fix(loop): strip input_audio and video_url before session persistence

Code review found that _sanitize_persisted_blocks only stripped
image_url blocks, causing base64-encoded audio/video payloads to
bloat session history files.

- Extend _sanitize_persisted_blocks to replace input_audio and
  video_url blocks with text placeholders using LLMProvider._media_placeholder.
- Add tests for audio/video stripping with and without _meta.path.

All 3267 tests pass.

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -49,7 +49,7 @@ body:
     attributes:
       label: nanobot Version
       description: Run `nanobot --version` or `pip show nanobot-ai`
-      placeholder: e.g., 0.1.5
+      placeholder: e.g., 0.2.0
     validations:
       required: true
 

Dockerfile

@@ -14,8 +14,9 @@ RUN apt-get update && \
 
 WORKDIR /app
 
-# Install Python dependencies first (cached layer)
-COPY pyproject.toml README.md LICENSE ./
+# Install Python dependencies first (cached layer). Hatch reads the custom build
+# hook from hatch_build.py even for this metadata-only install.
+COPY pyproject.toml README.md LICENSE THIRD_PARTY_NOTICES.md hatch_build.py ./
 RUN mkdir -p nanobot bridge && touch nanobot/__init__.py && \
     uv pip install --system --no-cache . && \
     rm -rf nanobot bridge
@@ -23,6 +24,7 @@ RUN mkdir -p nanobot bridge && touch nanobot/__init__.py && \
 # Copy the full source and install
 COPY nanobot/ nanobot/
 COPY bridge/ bridge/
+COPY webui/ webui/
 RUN uv pip install --system --no-cache .
 
 # Build the WhatsApp bridge
@@ -43,8 +45,8 @@ RUN sed -i 's/\r$//' /usr/local/bin/entrypoint.sh && chmod +x /usr/local/bin/ent
 USER nanobot
 ENV HOME=/home/nanobot
 
-# Gateway default port
-EXPOSE 18790
+# Gateway health endpoint and optional WebUI/WebSocket channel ports
+EXPOSE 18790 8765
 
 ENTRYPOINT ["entrypoint.sh"]
 CMD ["status"]

README.md

@@ -23,6 +23,7 @@
 
 ## 📢 News
 
+- **2026-05-15** 🚀 Released **v0.2.0** — **`/goal`** holds sustained objectives across turns, WebUI now ships inside the wheel, image generation end to end, 5 new providers with `fallback_models`, and a real agent-loop refactor. Please see [release notes](https://github.com/HKUDS/nanobot/releases/tag/v0.2.0) for details.
 - **2026-05-14** 🎯 **`/goal`** for long-term objectives, visible multi-step progress, long-horizon missions in chat.
 - **2026-05-13** 🧠 Streaming reasoning before answers, automatic backup models, smoother plug-in reconnects.
 - **2026-05-12** 🎛️ Saved model presets with WebUI badge, simpler plug-in tools, quieter Feishu topic threads.
@@ -211,13 +212,13 @@ nanobot agent
 
 
 - Want different LLM providers, web search, MCP, security settings, or more config options? See [Configuration](./docs/configuration.md)
+- Want to run locally? Use [Atomic Chat](./docs/configuration.md#atomic-chat-local), [vLLM](./docs/configuration.md#vllm-local-openai-compatible), [Ollama](./docs/configuration.md#ollama-local), and [others](./docs/configuration.md#local-providers).
 - Want to run nanobot in chat apps like Telegram, Discord, WeChat or Feishu? See [Chat Apps](./docs/chat-apps.md)
 - Want Docker or Linux service deployment? See [Deployment](./docs/deployment.md)
 
-## 🧪 WebUI (Development)
+## 🌐 WebUI
 
-> [!NOTE]
-> The WebUI development workflow currently requires a source checkout and is not yet shipped together with the official packaged release. See [WebUI Document](./webui/README.md) for full WebUI development docs and build steps.
+The WebUI ships **inside the published wheel** — no extra build step. Just enable the WebSocket channel and open it in your browser.
 
 <p align="center">
   <img src="images/nanobot_webui.png" alt="nanobot webui preview" width="900">
@@ -235,13 +236,12 @@ nanobot agent
 nanobot gateway
 ```
 
-**3. Start the webui dev server**
+**3. Open the WebUI**
 
-```bash
-cd webui
-bun install
-bun run dev
-```
+Visit [`http://127.0.0.1:8765`](http://127.0.0.1:8765) in your browser. To open it from another device on your LAN, see [WebUI docs → LAN access](./webui/README.md#access-from-another-device-lan).
+
+> [!TIP]
+> Working on the WebUI itself? Check out [`webui/README.md`](./webui/README.md) for the Vite dev server (HMR) workflow.
 
 ## 🏗️ Architecture
 
@@ -330,4 +330,4 @@ This project was started by [Xubin Ren](https://github.com/re-bin) as a personal
 <p align="center">
   <em> Thanks for visiting ✨ nanobot!</em><br><br>
   <img src="https://visitor-badge.laobi.icu/badge?page_id=HKUDS.nanobot&style=for-the-badge&color=00d4ff" alt="Views">
-</p>
+</p>

docker-compose.yml

@@ -20,6 +20,7 @@ services:
     restart: unless-stopped
     ports:
       - 18790:18790
+      - 8765:8765
     deploy:
       resources:
         limits:

docs/README.md

@@ -15,6 +15,7 @@ Start here for setup, everyday usage, and deployment.
 | Agent social network | [`agent-social-network.md`](./agent-social-network.md) | Join external agent communities from nanobot |
 | Configuration | [`configuration.md`](./configuration.md) | Providers, tools, channels, MCP, and runtime settings |
 | Image generation | [`image-generation.md`](./image-generation.md) | Configure image providers, WebUI image mode, and generated artifacts |
+| WebUI | [`../webui/README.md`](../webui/README.md) | Open the bundled browser UI; LAN access; Vite dev server for contributors |
 | Multiple instances | [`multiple-instances.md`](./multiple-instances.md) | Run isolated bots with separate configs and workspaces |
 | CLI reference | [`cli-reference.md`](./cli-reference.md) | Core CLI commands and common entrypoints |
 | In-chat commands | [`chat-commands.md`](./chat-commands.md) | Slash commands and periodic task behavior |

docs/configuration.md

@@ -26,7 +26,52 @@ Instead of storing secrets directly in `config.json`, you can use `${VAR_NAME}`
 }
 ```
 
-For **systemd** deployments, use `EnvironmentFile=` in the service unit to load variables from a file that only the deploying user can read:
+Any string value in `config.json` can use `${VAR_NAME}`. Resolution runs once at startup, in memory only — resolved values are never written back to disk, so editing config through `nanobot onboard` or the WebUI preserves the placeholder.
+
+If a referenced variable is unset, nanobot fails fast at startup with `ValueError: Environment variable 'NAME' referenced in config is not set`.
+
+### More examples
+
+**MCP servers** — both stdio `env` and HTTP `headers`:
+
+```json
+{
+  "tools": {
+    "mcpServers": {
+      "github": {
+        "command": "npx",
+        "args": ["-y", "@modelcontextprotocol/server-github"],
+        "env": { "GITHUB_PERSONAL_ACCESS_TOKEN": "${GITHUB_TOKEN}" }
+      },
+      "remote": {
+        "url": "https://example.com/mcp/",
+        "headers": { "Authorization": "Bearer ${REMOTE_MCP_TOKEN}" }
+      }
+    }
+  }
+}
+```
+
+**Web search providers:**
+
+```json
+{
+  "tools": {
+    "web": {
+      "search": {
+        "provider": "brave",
+        "apiKey": "${BRAVE_API_KEY}"
+      }
+    }
+  }
+}
+```
+
+### Loading variables at startup
+
+Pick whatever fits your deployment — nanobot only reads `os.environ` at startup, so any mechanism that populates the process environment works.
+
+**systemd** — use `EnvironmentFile=` in the service unit to load variables from a file that only the deploying user can read:
 
 ```ini
 # /etc/systemd/system/nanobot.service (excerpt)
@@ -42,6 +87,35 @@ TELEGRAM_TOKEN=your-token-here
 IMAP_PASSWORD=your-password-here
 ```
 
+**Docker** — pass an env file to the locally built image (one `KEY=VALUE` per line), or use `-e KEY=value`:
+
+```bash
+docker run --rm --env-file=./nanobot.env \
+  -v ~/.nanobot:/home/nanobot/.nanobot \
+  nanobot agent -m "Hello"
+```
+
+**direnv** — drop a `.envrc` in your working directory and run `direnv allow`:
+
+```bash
+# .envrc (auto-loaded by direnv)
+export TELEGRAM_TOKEN=your-token-here
+export ANTHROPIC_API_KEY=...
+```
+
+**Secret managers (1Password, Bitwarden, pass)** — wrap the process so secrets only exist as env vars for the lifetime of the run, never on disk:
+
+```bash
+# 1Password — references in .env.tpl look like `op://Vault/Item/field`
+op run --env-file=.env.tpl -- nanobot agent
+
+# pass (passwordstore.org)
+ANTHROPIC_API_KEY="$(pass show api/anthropic)" nanobot agent
+
+# Bitwarden
+ANTHROPIC_API_KEY="$(bw get password api/anthropic)" nanobot agent
+```
+
 ## Providers
 
 > [!TIP]
@@ -60,6 +134,7 @@ IMAP_PASSWORD=your-password-here
 | `custom` | Any OpenAI-compatible endpoint | — |
 | `openrouter` | LLM (recommended, access to all models) | [openrouter.ai](https://openrouter.ai) |
 | `huggingface` | LLM (Hugging Face Inference Providers) | [huggingface.co/settings/tokens](https://huggingface.co/settings/tokens) |
+| `skywork` | LLM (Skywork / APIFree API gateway) | [apifree.ai](https://www.apifree.ai) |
 | `volcengine` | LLM (VolcEngine, pay-per-use) | [Coding Plan](https://www.volcengine.com/activity/codingplan?utm_campaign=nanobot&utm_content=nanobot&utm_medium=devrel&utm_source=OWO&utm_term=nanobot) · [volcengine.com](https://www.volcengine.com) |
 | `byteplus` | LLM (VolcEngine international, pay-per-use) | [Coding Plan](https://www.byteplus.com/en/activity/codingplan?utm_campaign=nanobot&utm_content=nanobot&utm_medium=devrel&utm_source=OWO&utm_term=nanobot) · [byteplus.com](https://www.byteplus.com) |
 | `anthropic` | LLM (Claude direct) | [console.anthropic.com](https://console.anthropic.com) |
@@ -78,6 +153,7 @@ IMAP_PASSWORD=your-password-here
 | `zhipu` | LLM (Zhipu GLM) | [open.bigmodel.cn](https://open.bigmodel.cn) |
 | `mimo` | LLM (MiMo) | [platform.xiaomimimo.com](https://platform.xiaomimimo.com) |
 | `longcat` | LLM (LongCat) | [longcat.chat](https://longcat.chat/platform/docs/zh/) |
+| `ant_ling` | LLM (Ant Ling / 蚂蚁百灵) | [developer.ant-ling.com](https://developer.ant-ling.com/en/docs/api-reference/openai/) |
 | `ollama` | LLM (local, Ollama) | — |
 | `lm_studio` | LLM (local, LM Studio) | — |
 | `atomic_chat` | LLM (local, [Atomic Chat](https://atomic.chat/)) | — |
@@ -89,6 +165,36 @@ IMAP_PASSWORD=your-password-here
 | `github_copilot` | LLM (GitHub Copilot, OAuth) | `nanobot provider login github-copilot` |
 | `qianfan` | LLM (Baidu Qianfan) | [cloud.baidu.com](https://cloud.baidu.com/doc/qianfan/s/Hmh4suq26) |
 
+<details>
+<summary><b>Skywork / APIFree</b></summary>
+
+Skywork uses the OpenAI-compatible APIFree API endpoint. Configure the provider
+once, then use Skywork model IDs such as `skywork-ai/skyclaw-v1`.
+
+```json
+{
+  "providers": {
+    "skywork": {
+      "apiKey": "${SKYWORK_API_KEY}",
+      "apiBase": "https://api.apifree.ai/v1"
+    }
+  },
+  "agents": {
+    "defaults": {
+      "provider": "skywork",
+      "model": "skywork-ai/skyclaw-v1",
+      "maxTokens": 32768,
+      "contextWindowTokens": 131072
+    }
+  }
+}
+```
+
+You can also reference `${APIFREE_API_KEY}` in `apiKey` if that is how your
+environment names the credential.
+
+</details>
+
 <details>
 <summary><b>AWS Bedrock (Converse API)</b></summary>
 
@@ -370,6 +476,34 @@ Official model names include `LongCat-Flash-Chat`, `LongCat-Flash-Thinking`,
 
 </details>
 
+<details>
+<summary><b>Ant Ling (OpenAI-compatible)</b></summary>
+
+Ant Ling is available through nanobot's built-in OpenAI-compatible provider flow.
+The default API base points to `https://api.ant-ling.com/v1`, so you usually
+only need to set `apiKey`.
+
+```json
+{
+  "providers": {
+    "antLing": {
+      "apiKey": "${ANT_LING_API_KEY}"
+    }
+  },
+  "agents": {
+    "defaults": {
+      "provider": "ant_ling",
+      "model": "Ling-2.6-flash"
+    }
+  }
+}
+```
+
+Official OpenAI-compatible model names include `Ling-2.6-1T`,
+`Ling-2.6-flash`, `Ling-2.5-1T`, `Ling-1T`, `Ring-2.5-1T`, and `Ring-1T`.
+
+</details>
+
 <details>
 <summary><b>Custom Provider (Any OpenAI-compatible API)</b></summary>
 
@@ -438,6 +572,8 @@ Some OpenAI-compatible gateways expose request-body extensions such as vLLM guid
 
 </details>
 
+<a id="local-providers"></a>
+<a id="ollama-local"></a>
 <details>
 <summary><b>Ollama (local)</b></summary>
 
@@ -503,12 +639,19 @@ ollama run llama3.2
 
 </details>
 
+<a id="atomic-chat-local"></a>
 <details>
 <summary><b>Atomic Chat (local)</b></summary>
 
-[Atomic Chat](https://atomic.chat/) is a local-first desktop app that exposes an **OpenAI-compatible** HTTP API (default `http://localhost:1337/v1`). Start Atomic Chat and enable the local API server, then point nanobot at it.
+[Atomic Chat](https://atomic.chat/) is a local-first desktop app that exposes an **OpenAI-compatible** HTTP API (default `http://localhost:1337/v1`). Use it when you want to run nanobot against a model on your own machine instead of a hosted API provider.
 
-**1. Add to config** (partial — merge into `~/.nanobot/config.json`):
+**1. Start Atomic Chat**
+
+- Install [Atomic Chat](https://atomic.chat/) on your machine.
+- Open Atomic Chat, download a model, and keep the app running. The local API is enabled by default.
+- Copy the model ID exposed by the local API. For example, the model ID for `Qwen 3 32B` might be `qwen3-32b`.
+
+**2. Add to config** (partial — merge into `~/.nanobot/config.json`):
 
 ```json
 {
@@ -521,13 +664,13 @@ ollama run llama3.2
   "agents": {
     "defaults": {
       "provider": "atomic_chat",
-      "model": "your-model-id-from-atomic-chat"
+      "model": "qwen3-32b"
     }
   }
 }
 ```
 
-> **Note:** Set `apiKey` to `null` if your Atomic Chat server does not require a key. If it does, set `apiKey` (or the `ATOMIC_CHAT_API_KEY` environment variable) to the value Atomic Chat expects. The `model` string must match the model id Atomic Chat exposes on its OpenAI-compatible endpoint.
+> **Note:** Replace `qwen3-32b` with the model ID from Atomic Chat. Set `apiKey` to `null` if your Atomic Chat server does not require a key. If it does, set `apiKey` (or the `ATOMIC_CHAT_API_KEY` environment variable) to the value Atomic Chat expects.
 
 > `provider: "auto"` also works when `providers.atomic_chat.apiBase` is configured, but setting `"provider": "atomic_chat"` is the clearest option.
 
@@ -608,6 +751,7 @@ docker run -d \
 > See the [official OVMS docs](https://docs.openvino.ai/2026/model-server/ovms_docs_llm_quickstart.html) for more details.
 </details>
 
+<a id="vllm-local-openai-compatible"></a>
 <details>
 <summary><b>vLLM (local / OpenAI-compatible)</b></summary>
 
@@ -917,7 +1061,7 @@ By default, web search uses `duckduckgo`, and it works out of the box without an
     "web": {
       "search": {
         "provider": "brave",
-        "apiKey": "BSA..."
+        "apiKey": "${BRAVE_API_KEY}"
       }
     }
   }
@@ -931,7 +1075,7 @@ By default, web search uses `duckduckgo`, and it works out of the box without an
     "web": {
       "search": {
         "provider": "tavily",
-        "apiKey": "tvly-..."
+        "apiKey": "${TAVILY_API_KEY}"
       }
     }
   }
@@ -945,7 +1089,7 @@ By default, web search uses `duckduckgo`, and it works out of the box without an
     "web": {
       "search": {
         "provider": "jina",
-        "apiKey": "jina_..."
+        "apiKey": "${JINA_API_KEY}"
       }
     }
   }
@@ -959,7 +1103,7 @@ By default, web search uses `duckduckgo`, and it works out of the box without an
     "web": {
       "search": {
         "provider": "kagi",
-        "apiKey": "your-kagi-api-key"
+        "apiKey": "${KAGI_API_KEY}"
       }
     }
   }
@@ -973,7 +1117,7 @@ By default, web search uses `duckduckgo`, and it works out of the box without an
     "web": {
       "search": {
         "provider": "olostep",
-        "apiKey": "YOUR_OLOSTEP_API_KEY"
+        "apiKey": "${OLOSTEP_API_KEY}"
       }
     }
   }
@@ -1136,6 +1280,8 @@ MCP tools are automatically discovered and registered on startup. The LLM can us
 > [!TIP]
 > For production deployments, set `"restrictToWorkspace": true` and `"tools.exec.sandbox": "bwrap"` in your config to sandbox the agent.
 
+For API keys, tokens, and other secrets, see [Environment Variables for Secrets](#environment-variables-for-secrets) — avoid storing them directly in `config.json`.
+
 | Option | Default | Description |
 |--------|---------|-------------|
 | `tools.restrictToWorkspace` | `false` | When `true`, restricts **all** agent tools (shell, file read/write/edit, list) to the workspace directory. Prevents path traversal and out-of-scope access. |

docs/deployment.md

@@ -10,6 +10,18 @@
 > [!IMPORTANT]
 > Official Docker usage currently means building from this repository with the included `Dockerfile`. Docker Hub images under third-party namespaces are not maintained or verified by HKUDS/nanobot; do not mount API keys or bot tokens into them unless you trust the publisher.
 
+> [!IMPORTANT]
+> The gateway and WebSocket channel default to `host: "127.0.0.1"` in `config.json` (set in `nanobot/config/schema.py`). Docker `-p` port forwarding cannot reach a container's loopback interface, so for the host or LAN to reach the exposed ports you must set both binds to `0.0.0.0` in `~/.nanobot/config.json` before starting the container:
+>
+> ```json
+> {
+>   "gateway":  { "host": "0.0.0.0" },
+>   "channels": { "websocket": { "host": "0.0.0.0" } }
+> }
+> ```
+>
+> When `host` is `0.0.0.0`, the gateway refuses to start unless `token` or `tokenIssueSecret` is also configured on the WebSocket channel — see [`webui/README.md`](../webui/README.md) for details.
+
 ### Docker Compose
 
 ```bash
@@ -36,8 +48,20 @@ docker run -v ~/.nanobot:/home/nanobot/.nanobot --rm nanobot onboard
 # Edit config on host to add API keys
 vim ~/.nanobot/config.json
 
-# Run gateway (connects to enabled channels, e.g. Telegram/Discord/Mochat)
-docker run -v ~/.nanobot:/home/nanobot/.nanobot -p 18790:18790 nanobot gateway
+# Run gateway (connects to enabled channels, e.g. Telegram/Discord/Mochat).
+# Mirrors the security caps and port mappings declared in docker-compose.yml:
+#   - `--cap-drop ALL --cap-add SYS_ADMIN` + unconfined apparmor/seccomp are required
+#     when `tools.exec.sandbox: "bwrap"` is enabled (bwrap needs CAP_SYS_ADMIN for
+#     user namespaces). Without them, `bwrap` exits with `clone3: Operation not permitted`.
+#   - `-p 8765:8765` exposes the WebSocket channel / WebUI alongside the gateway health
+#     endpoint on 18790.
+docker run \
+  --cap-drop ALL --cap-add SYS_ADMIN \
+  --security-opt apparmor=unconfined \
+  --security-opt seccomp=unconfined \
+  -v ~/.nanobot:/home/nanobot/.nanobot \
+  -p 18790:18790 -p 8765:8765 \
+  nanobot gateway
 
 # Or run a single command
 docker run -v ~/.nanobot:/home/nanobot/.nanobot --rm nanobot agent -m "Hello!"

docs/image-generation.md

@@ -6,8 +6,6 @@ The feature is disabled by default. Enable it in `~/.nanobot/config.json`, confi
 
 ## Quick Setup
 
-OpenRouter example:
-
 ```json
 {
   "providers": {
@@ -19,34 +17,13 @@ OpenRouter example:
     "imageGeneration": {
       "enabled": true,
       "provider": "openrouter",
-      "model": "openai/gpt-5.4-image-2",
-      "defaultAspectRatio": "1:1",
-      "defaultImageSize": "1K"
+      "model": "openai/gpt-5.4-image-2"
     }
   }
 }
 ```
 
-AIHubMix example:
-
-```json
-{
-  "providers": {
-    "aihubmix": {
-      "apiKey": "${AIHUBMIX_API_KEY}"
-    }
-  },
-  "tools": {
-    "imageGeneration": {
-      "enabled": true,
-      "provider": "aihubmix",
-      "model": "gpt-image-2-free",
-      "defaultAspectRatio": "1:1",
-      "defaultImageSize": "1K"
-    }
-  }
-}
-```
+See [Provider Notes](#provider-notes) for AIHubMix, MiniMax, and Gemini configuration examples.
 
 > [!TIP]
 > Prefer environment variables for API keys. nanobot resolves `${VAR_NAME}` values from the environment at startup.
@@ -69,7 +46,7 @@ The WebUI hides provider storage details from the user. The agent sees the saved
 | Option | Type | Default | Description |
 |--------|------|---------|-------------|
 | `tools.imageGeneration.enabled` | boolean | `false` | Register the `generate_image` tool |
-| `tools.imageGeneration.provider` | string | `"openrouter"` | Image provider name. Currently `openrouter` and `aihubmix` are supported |
+| `tools.imageGeneration.provider` | string | `"openrouter"` | Image provider name. Supported values: `openrouter`, `aihubmix`, `minimax`, `gemini`, `stepfun` |
 | `tools.imageGeneration.model` | string | `"openai/gpt-5.4-image-2"` | Provider model name |
 | `tools.imageGeneration.defaultAspectRatio` | string | `"1:1"` | Default ratio when the prompt/tool call does not specify one |
 | `tools.imageGeneration.defaultImageSize` | string | `"1K"` | Default size hint, for example `1K`, `2K`, `4K`, or `1024x1024` |
@@ -139,6 +116,110 @@ Configure:
 
 `quality: low` is optional. It can make free image models faster and less likely to time out, but it is not required for correctness.
 
+### MiniMax
+
+MiniMax `image-01` supports text-to-image and reference-image (subject reference) edits. Supported aspect ratios are `1:1`, `16:9`, `4:3`, `3:2`, `2:3`, `3:4`, `9:16`, and `21:9`.
+
+```json
+{
+  "providers": {
+    "minimax": {
+      "apiKey": "${MINIMAX_API_KEY}"
+    }
+  },
+  "tools": {
+    "imageGeneration": {
+      "enabled": true,
+      "provider": "minimax",
+      "model": "image-01",
+      "defaultAspectRatio": "1:1"
+    }
+  }
+}
+```
+
+### Gemini
+
+nanobot supports two Gemini image generation model families via Google's Generative Language API:
+
+| Model | Endpoint | Reference images |
+|-------|----------|-----------------|
+| `imagen-4.0-generate-001` | `:predict` | Not supported by this integration |
+| `gemini-2.5-flash-image` | `:generateContent` | Supported |
+
+For reference-image edits, use a Gemini Flash image model:
+
+```json
+{
+  "providers": {
+    "gemini": {
+      "apiKey": "${GEMINI_API_KEY}"
+    }
+  },
+  "tools": {
+    "imageGeneration": {
+      "enabled": true,
+      "provider": "gemini",
+      "model": "gemini-2.5-flash-image"
+    }
+  }
+}
+```
+
+Imagen 4 supports the aspect ratios `1:1`, `9:16`, `16:9`, `3:4`, and `4:3`. Unsupported ratios are ignored and the model uses its default. The `defaultImageSize` setting has no effect on Gemini models; sizing is controlled by `defaultAspectRatio` only. Reference images passed with an Imagen model are ignored (with a warning logged).
+
+### StepFun
+
+StepFun (阶跃星辰) `step-image-edit-2` supports text-to-image generation.  The `step-1x-medium` variant additionally supports **style-reference** image edits, where a reference image guides the visual style of the output.
+
+Supported aspect ratios: `1:1`, `16:9`, `9:16`, `3:4`, `4:3`.  Sizes are specified as `WIDTHxHEIGHT` (e.g. `1024x1024`, `1280x800`, `800x1280`).
+
+```json
+{
+  "providers": {
+    "stepfun": {
+      "apiKey": "${STEPFUN_API_KEY}"
+    }
+  },
+  "tools": {
+    "imageGeneration": {
+      "enabled": true,
+      "provider": "stepfun",
+      "model": "step-image-edit-2"
+    }
+  }
+}
+```
+
+> [!NOTE]
+> The StepFun provider reuses the existing `providers.stepfun` config block (the same one used for StepFun's LLM API).  Set `providers.stepfun.apiKey` once and it is shared between text and image generation.
+>
+> When `step-image-edit-2` is used, `reference_images` are ignored (the model does not support style reference).  Switch to `step-1x-medium` to use reference-image-guided generation.
+
+#### StepPlan (Subscription)
+
+StepPlan is StepFun's subscription tier and uses a different API base URL. The image generation endpoint path is the same — just override `apiBase`:
+
+```json
+{
+  "providers": {
+    "stepfun": {
+      "apiKey": "${STEPFUN_API_KEY}",
+      "apiBase": "https://api.stepfun.com/step_plan/v1"
+    }
+  },
+  "tools": {
+    "imageGeneration": {
+      "enabled": true,
+      "provider": "stepfun",
+      "model": "step-image-edit-2"
+    }
+  }
+}
+```
+
+`apiBase` takes precedence over the registry default, so with the StepPlan base URL configured, image requests are sent to `https://api.stepfun.com/step_plan/v1/images/generations` — the same path prefix used for LLM calls. The API key is shared with the standard StepFun provider.
+
 ## Artifacts
 
 Generated images are stored under the active nanobot instance's media directory:
@@ -193,7 +274,7 @@ Use the reference image. Keep the same robot and composition, change the palette
 |---------|-------|
 | `generate_image` is not available | Set `tools.imageGeneration.enabled` to `true` and restart the gateway |
 | Missing API key error | Configure `providers.<provider>.apiKey`; if using `${VAR_NAME}`, confirm the environment variable is visible to the gateway process |
-| `unsupported image generation provider` | Use `openrouter` or `aihubmix` |
+| `unsupported image generation provider` | Use `openrouter`, `aihubmix`, `minimax`, `gemini`, or `stepfun` |
 | AIHubMix says `Incorrect model ID` | Use `model: "gpt-image-2-free"`; nanobot expands it to the required `openai/gpt-image-2-free` model path internally |
 | Generation times out | Try a smaller/default image size, set AIHubMix `extraBody.quality` to `"low"`, or retry later |
 | Reference image rejected | Reference image paths must be inside the workspace or nanobot media directory and must be valid image files |

hatch_build.py

@@ -0,0 +1,101 @@
+"""Hatch build hook that bundles the webui (Vite) into nanobot/web/dist.
+
+Triggered automatically by `python -m build` (and any other hatch-driven build)
+so published wheels and sdists ship a fresh webui without requiring developers
+to remember `cd webui && bun run build` beforehand.
+
+Behaviour:
+
+- Skips for editable installs (`pip install -e .`). Editable mode is for Python
+  development; webui contributors use `cd webui && bun run dev` (Vite HMR) and
+  do not need a packaged `dist/`.
+- No-op when `webui/package.json` is absent (e.g. installing from an sdist that
+  already contains a prebuilt `nanobot/web/dist/`).
+- Skips when `NANOBOT_SKIP_WEBUI_BUILD=1` is set.
+- Skips when `nanobot/web/dist/index.html` already exists, unless
+  `NANOBOT_FORCE_WEBUI_BUILD=1` is set.
+- Uses `bun` when available, otherwise falls back to `npm`. The chosen tool
+  performs `install` followed by `run build`.
+"""
+
+from __future__ import annotations
+
+import os
+import shutil
+import subprocess
+from pathlib import Path
+
+from hatchling.builders.hooks.plugin.interface import BuildHookInterface
+
+
+class WebUIBuildHook(BuildHookInterface):
+    PLUGIN_NAME = "webui-build"
+
+    def initialize(self, version: str, build_data: dict) -> None:  # noqa: D401
+        root = Path(self.root)
+        webui_dir = root / "webui"
+        package_json = webui_dir / "package.json"
+        dist_dir = root / "nanobot" / "web" / "dist"
+        index_html = dist_dir / "index.html"
+
+        # `pip install -e .` builds an editable wheel; skip the (slow) webui
+        # bundle since editable installs target Python development and webui
+        # work uses `bun run dev` instead.
+        if self.target_name == "wheel" and version == "editable":
+            self.app.display_info(
+                "[webui-build] skipped for editable install "
+                "(use `cd webui && bun run build` to bundle webui manually)"
+            )
+            return
+
+        if os.environ.get("NANOBOT_SKIP_WEBUI_BUILD") == "1":
+            self.app.display_info("[webui-build] skipped via NANOBOT_SKIP_WEBUI_BUILD=1")
+            return
+
+        if not package_json.is_file():
+            self.app.display_info(
+                "[webui-build] no webui/ source tree, assuming prebuilt nanobot/web/dist/"
+            )
+            return
+
+        force = os.environ.get("NANOBOT_FORCE_WEBUI_BUILD") == "1"
+        if index_html.is_file() and not force:
+            self.app.display_info(
+                f"[webui-build] reusing existing build at {dist_dir} "
+                "(set NANOBOT_FORCE_WEBUI_BUILD=1 to rebuild)"
+            )
+            return
+
+        runner = self._pick_runner()
+        if runner is None:
+            raise RuntimeError(
+                "[webui-build] neither `bun` nor `npm` is available on PATH; "
+                "install one or set NANOBOT_SKIP_WEBUI_BUILD=1 to bypass."
+            )
+
+        self.app.display_info(f"[webui-build] using {runner} to build webui")
+        self._run([runner, "install"], cwd=webui_dir)
+        self._run([runner, "run", "build"], cwd=webui_dir)
+
+        if not index_html.is_file():
+            raise RuntimeError(
+                f"[webui-build] build finished but {index_html} is missing; "
+                "check webui/vite.config.ts outDir."
+            )
+        self.app.display_info(f"[webui-build] webui ready at {dist_dir}")
+
+    @staticmethod
+    def _pick_runner() -> str | None:
+        for candidate in ("bun", "npm"):
+            if shutil.which(candidate):
+                return candidate
+        return None
+
+    def _run(self, cmd: list[str], *, cwd: Path) -> None:
+        self.app.display_info(f"[webui-build] $ {' '.join(cmd)} (cwd={cwd})")
+        try:
+            subprocess.run(cmd, cwd=cwd, check=True)
+        except subprocess.CalledProcessError as exc:
+            raise RuntimeError(
+                f"[webui-build] command failed ({exc.returncode}): {' '.join(cmd)}"
+            ) from exc

nanobot/__init__.py

@@ -21,7 +21,7 @@ def _resolve_version() -> str:
         return _pkg_version("nanobot-ai")
     except PackageNotFoundError:
         # Source checkouts often import nanobot without installed dist-info.
-        return _read_pyproject_version() or "0.1.5.post3"
+        return _read_pyproject_version() or "0.2.0"
 
 
 __version__ = _resolve_version()

nanobot/agent/autocompact.py

@@ -4,7 +4,7 @@ from __future__ import annotations
 
 from collections.abc import Collection
 from datetime import datetime
-from typing import TYPE_CHECKING, Any, Callable, Coroutine
+from typing import TYPE_CHECKING, Callable, Coroutine
 
 from loguru import logger
 
@@ -37,27 +37,6 @@ class AutoCompact:
     def _format_summary(text: str, last_active: datetime) -> str:
         return f"Previous conversation summary (last active {last_active.isoformat()}):\n{text}"
 
-    def _split_unconsolidated(
-        self, session: Session,
-    ) -> tuple[list[dict[str, Any]], list[dict[str, Any]]]:
-        """Split live session tail into archiveable prefix and retained recent suffix."""
-        tail = list(session.messages[session.last_consolidated:])
-        if not tail:
-            return [], []
-
-        probe = Session(
-            key=session.key,
-            messages=tail.copy(),
-            created_at=session.created_at,
-            updated_at=session.updated_at,
-            metadata={},
-            last_consolidated=0,
-        )
-        probe.retain_recent_legal_suffix(self._RECENT_SUFFIX_MESSAGES)
-        kept = probe.messages
-        cut = len(tail) - len(kept)
-        return tail[:cut], kept
-
     def check_expired(self, schedule_background: Callable[[Coroutine], None],
                       active_session_keys: Collection[str] = ()) -> None:
         """Schedule archival for idle sessions, skipping those with in-flight agent tasks."""
@@ -74,33 +53,17 @@ class AutoCompact:
 
     async def _archive(self, key: str) -> None:
         try:
-            self.sessions.invalidate(key)
-            session = self.sessions.get_or_create(key)
-            archive_msgs, kept_msgs = self._split_unconsolidated(session)
-            if not archive_msgs and not kept_msgs:
-                session.updated_at = datetime.now()
-                self.sessions.save(session)
-                return
-
-            last_active = session.updated_at
-            summary = ""
-            if archive_msgs:
-                summary = await self.consolidator.archive(archive_msgs) or ""
+            summary = await self.consolidator.compact_idle_session(
+                key, self._RECENT_SUFFIX_MESSAGES,
+            )
             if summary and summary != "(nothing)":
-                self._summaries[key] = (summary, last_active)
-                session.metadata["_last_summary"] = {"text": summary, "last_active": last_active.isoformat()}
-            session.messages = kept_msgs
-            session.last_consolidated = 0
-            session.updated_at = datetime.now()
-            self.sessions.save(session)
-            if archive_msgs:
-                logger.info(
-                    "Auto-compact: archived {} (archived={}, kept={}, summary={})",
-                    key,
-                    len(archive_msgs),
-                    len(kept_msgs),
-                    bool(summary),
-                )
+                session = self.sessions.get_or_create(key)
+                meta = session.metadata.get("_last_summary")
+                if isinstance(meta, dict):
+                    self._summaries[key] = (
+                        meta["text"],
+                        datetime.fromisoformat(meta["last_active"]),
+                    )
         except Exception:
             logger.exception("Auto-compact: failed for {}", key)
         finally:

nanobot/agent/context.py

@@ -2,6 +2,7 @@
 
 import base64
 import mimetypes
+import os
 import platform
 from contextlib import suppress
 from importlib.resources import files as pkg_files
@@ -10,11 +11,16 @@ from typing import Any, Mapping, Sequence
 
 from nanobot.agent.memory import MemoryStore
 from nanobot.agent.skills import SkillsLoader
+from nanobot.config.schema import InputLimitsConfig
 from nanobot.session.goal_state import goal_state_runtime_lines
 from nanobot.utils.helpers import (
+    audio_format_for_api,
+    audio_mime_compat,
     current_time_str,
+    detect_audio_mime,
     detect_image_mime,
     truncate_text,
+    video_mime_compat,
 )
 from nanobot.utils.prompt_templates import render_template
 
@@ -28,11 +34,12 @@ class ContextBuilder:
     _MAX_HISTORY_CHARS = 32_000  # hard cap on recent history section size
     _RUNTIME_CONTEXT_END = "[/Runtime Context]"
 
-    def __init__(self, workspace: Path, timezone: str | None = None, disabled_skills: list[str] | None = None):
+    def __init__(self, workspace: Path, timezone: str | None = None, disabled_skills: list[str] | None = None, input_limits: InputLimitsConfig | None = None):
         self.workspace = workspace
         self.timezone = timezone
         self.memory = MemoryStore(workspace)
         self.skills = SkillsLoader(workspace, disabled_skills=set(disabled_skills) if disabled_skills else None)
+        self.input_limits = input_limits or InputLimitsConfig()
 
     def build_system_prompt(
         self,
@@ -142,6 +149,28 @@ class ContextBuilder:
                 return content.strip() == tpl.read_text(encoding="utf-8").strip()
         return False
 
+    @staticmethod
+    def _file_size_ok(p: Path, max_bytes: int) -> bool | None:
+        """Check file size via stat without reading into memory.
+
+        Returns True if size is within limit, False if oversized,
+        None if file cannot be stat'd (caller should try read_bytes instead).
+        """
+        try:
+            return os.stat(p).st_size <= max_bytes
+        except OSError:
+            return None
+
+    @staticmethod
+    def _encode_image_block(raw: bytes, mime: str, path: Path) -> dict[str, Any]:
+        """Base64-encode file bytes into an image_url content block."""
+        b64 = base64.b64encode(raw).decode()
+        return {
+            "type": "image_url",
+            "image_url": {"url": f"data:{mime};base64,{b64}"},
+            "_meta": {"path": str(path)},
+        }
+
     def build_messages(
         self,
         history: list[dict[str, Any]],
@@ -154,6 +183,9 @@ class ContextBuilder:
         sender_id: str | None = None,
         session_summary: str | None = None,
         session_metadata: Mapping[str, Any] | None = None,
+        supports_vision: bool | None = None,
+        supports_audio: bool | None = None,
+        supports_video: bool | None = None,
     ) -> list[dict[str, Any]]:
         """Build the complete message list for an LLM call."""
         extra = goal_state_runtime_lines(session_metadata)
@@ -164,7 +196,12 @@ class ContextBuilder:
             sender_id=sender_id,
             supplemental_lines=extra or None,
         )
-        user_content = self._build_user_content(current_message, media)
+        user_content = self._build_user_content(
+            current_message, media,
+            supports_vision=supports_vision,
+            supports_audio=supports_audio,
+            supports_video=supports_video,
+        )
 
         # Merge runtime context and user content into a single user message
         # to avoid consecutive same-role messages that some providers reject.
@@ -186,28 +223,171 @@ class ContextBuilder:
         messages.append({"role": current_role, "content": merged})
         return messages
 
-    def _build_user_content(self, text: str, media: list[str] | None) -> str | list[dict[str, Any]]:
-        """Build user message content with optional base64-encoded images."""
+    def _build_user_content(
+        self,
+        text: str,
+        media: list[str] | None,
+        *,
+        supports_vision: bool | None = None,
+        supports_audio: bool | None = None,
+        supports_video: bool | None = None,
+    ) -> str | list[dict[str, Any]]:
+        """Build user message content with optional media blocks.
+
+        Args:
+            text: The user text message.
+            media: List of file paths to media files.
+            supports_vision: True=model supports images, False=use placeholder,
+                             None=unconfigured (send images as before, let
+                             provider/retry handle degradation).
+            supports_audio: True=model supports native audio, False/None=skip
+                            (channel layer already transcribed).
+            supports_video: True=model supports native video, False/None=use
+                            [file: path] placeholder.
+        """
         if not media:
             return text
 
-        images = []
+        blocks: list[dict[str, Any]] = []
+        notes: list[str] = []
+        limits = self.input_limits
+
+        # Enforce image count limit
+        max_images = limits.max_input_images
+        image_count = 0
+        image_media = []
+        non_image_media = []
         for path in media:
+            p = Path(path)
+            guessed_mime = mimetypes.guess_type(path)[0] or ""
+            if guessed_mime.startswith("image/"):
+                image_count += 1
+                if image_count <= max_images:
+                    image_media.append(path)
+            else:
+                non_image_media.append(path)
+
+        if image_count > max_images:
+            extra = image_count - max_images
+            noun = "image" if extra == 1 else "images"
+            notes.append(
+                f"[Skipped {extra} {noun}: "
+                f"only the first {max_images} images are included]"
+            )
+
+        # Process images
+        for path in image_media:
             p = Path(path)
             if not p.is_file():
                 continue
-            raw = p.read_bytes()
-            mime = detect_image_mime(raw) or mimetypes.guess_type(path)[0]
-            if not mime or not mime.startswith("image/"):
+
+            # When explicitly marked as non-vision, downgrade to text placeholder
+            if supports_vision is False:
+                blocks.append({"type": "text", "text": f"[image: {p}]"})
                 continue
-            b64 = base64.b64encode(raw).decode()
-            images.append({
-                "type": "image_url",
-                "image_url": {"url": f"data:{mime};base64,{b64}"},
-                "_meta": {"path": str(p)},
-            })
 
-        if not images:
-            return text
-        return images + [{"type": "text", "text": text}]
+            size_ok = self._file_size_ok(p, limits.max_input_image_bytes)
+            if size_ok is False:
+                size_mb = limits.max_input_image_bytes // (1024 * 1024)
+                notes.append(f"[Skipped image: file too large ({p.name}, limit {size_mb} MB)]")
+                continue
+            try:
+                raw = p.read_bytes()
+            except OSError:
+                notes.append(f"[Skipped image: unable to read ({p.name or path})]")
+                continue
+            img_mime = detect_image_mime(raw[:32]) or mimetypes.guess_type(path)[0]
+            if not img_mime or not img_mime.startswith("image/"):
+                notes.append(f"[Skipped image: unsupported or invalid image format ({p.name})]")
+                continue
+            blocks.append(self._encode_image_block(raw, img_mime, p))
+
+        # Process non-image media (audio, video, unknown)
+        audio_count = 0
+        video_count = 0
+        for path in non_image_media:
+            p = Path(path)
+            if not p.is_file():
+                continue
+            guessed_mime = mimetypes.guess_type(path)[0] or ""
+            is_audio = guessed_mime.startswith("audio/")
+            is_video = guessed_mime.startswith("video/")
+
+            # Pre-check file size via stat to avoid reading oversized files into memory.
+            # Determine the relevant byte limit based on detected media type.
+            _size_limit = 0
+            if is_audio or is_video:
+                _size_limit = limits.max_input_audio_bytes if is_audio else limits.max_input_video_bytes
+            _stat_size_ok = self._file_size_ok(p, _size_limit) if _size_limit else None
+            if _stat_size_ok is False:
+                size_mb = _size_limit // (1024 * 1024)
+                label = "audio" if is_audio else "video"
+                notes.append(f"[Skipped {label}: file too large ({p.name}, limit {size_mb} MB)]")
+                continue
+
+            try:
+                raw = p.read_bytes()
+            except OSError:
+                notes.append(f"[Skipped file: unable to read ({p.name or path})]")
+                continue
+
+            # Audio detection: by magic bytes or by filename
+            # Always pass filename so fallback can match when magic bytes fail
+            audio_mime = detect_audio_mime(raw[:32], filename=path)
+            if audio_mime or is_audio:
+                if supports_audio is True and audio_mime_compat(audio_mime):
+                    audio_count += 1
+                    if audio_count > limits.max_input_audios:
+                        if audio_count == limits.max_input_audios + 1:
+                            notes.append(
+                                f"[Skipped audio: only {limits.max_input_audios} audio file(s) allowed]"
+                            )
+                        continue
+                    if len(raw) > limits.max_input_audio_bytes:
+                        size_mb = limits.max_input_audio_bytes // (1024 * 1024)
+                        notes.append(f"[Skipped audio: file too large ({p.name}, limit {size_mb} MB)]")
+                        continue
+                    b64 = base64.b64encode(raw).decode()
+                    blocks.append({
+                        "type": "input_audio",
+                        "input_audio": {"data": b64, "format": audio_format_for_api(audio_mime)},
+                        "_meta": {"path": str(p)},
+                    })
+                else:
+                    blocks.append({"type": "text", "text": f"[audio: {p}]"})
+                continue
+
+            # Video detection (already classified above)
+            if is_video:
+                if supports_video is True and video_mime_compat(guessed_mime):
+                    video_count += 1
+                    if video_count > limits.max_input_videos:
+                        if video_count == limits.max_input_videos + 1:
+                            notes.append(
+                                f"[Skipped video: only {limits.max_input_videos} video file(s) allowed]"
+                            )
+                        continue
+                    if len(raw) > limits.max_input_video_bytes:
+                        size_mb = limits.max_input_video_bytes // (1024 * 1024)
+                        notes.append(f"[Skipped video: file too large ({p.name}, limit {size_mb} MB)]")
+                        continue
+                    b64 = base64.b64encode(raw).decode()
+                    blocks.append({
+                        "type": "video_url",
+                        "video_url": {"url": f"data:{guessed_mime};base64,{b64}"},
+                        "_meta": {"path": str(p)},
+                    })
+                else:
+                    blocks.append({"type": "text", "text": f"[video: {p}]"})
+                continue
+
+            # Unknown files are silently ignored (preserves pre-multimodal behaviour)
+            continue
+
+        note_text = "\n".join(notes).strip()
+        text_block = text if not note_text else (f"{note_text}\n\n{text}" if text else note_text)
+
+        if not blocks:
+            return text_block
+        return blocks + [{"type": "text", "text": text_block}]
 

nanobot/agent/loop.py

@@ -33,19 +33,20 @@ from nanobot.config.schema import AgentDefaults, ModelPresetConfig
 from nanobot.providers.base import LLMProvider
 from nanobot.providers.factory import ProviderSnapshot
 from nanobot.session.goal_state import (
-    goal_state_ws_blob,
     runner_wall_llm_timeout_s,
 )
 from nanobot.session.manager import Session, SessionManager
-from nanobot.utils.artifacts import generated_image_paths_from_messages
+from nanobot.session.webui_turns import (
+    WebuiTurnCoordinator,
+    build_bus_progress_callback,
+    mark_webui_session,
+)
 from nanobot.utils.document import extract_documents
 from nanobot.utils.helpers import image_placeholder_text
 from nanobot.utils.helpers import truncate_text as truncate_text_fn
 from nanobot.utils.image_generation_intent import image_generation_prompt
+from nanobot.utils.llm_runtime import LLMRuntime
 from nanobot.utils.runtime import EMPTY_FINAL_RESPONSE_MESSAGE
-from nanobot.utils.session_attachments import merge_turn_media_into_last_assistant
-from nanobot.utils.webui_titles import mark_webui_session, maybe_generate_webui_title_after_turn
-from nanobot.utils.webui_turn_helpers import publish_turn_run_status
 
 if TYPE_CHECKING:
     from nanobot.config.schema import (
@@ -100,7 +101,6 @@ class TurnContext:
     save_skip: int = 0
 
     outbound: OutboundMessage | None = None
-    generated_media: list[str] = field(default_factory=list)
 
     on_progress: Callable[..., Awaitable[None]] | None = None
     on_stream: Callable[[str], Awaitable[None]] | None = None
@@ -136,6 +136,11 @@ class AgentLoop:
     def tool_names(self) -> list[str]:
         return self.tools.tool_names
 
+    def llm_runtime(self) -> LLMRuntime:
+        """Return the current provider/model pair owned by this loop."""
+        self._refresh_provider_snapshot()
+        return LLMRuntime(self.provider, self.model)
+
     _RUNTIME_CHECKPOINT_KEY = "runtime_checkpoint"
     _PENDING_USER_TURN_KEY = "pending_user_turn"
 
@@ -185,6 +190,10 @@ class AgentLoop:
         model_preset: str | None = None,
         preset_snapshot_loader: preset_helpers.PresetSnapshotLoader | None = None,
         runtime_model_publisher: Callable[[str, str | None], None] | None = None,
+        input_limits: Any = None,
+        supports_vision: bool | None = None,
+        supports_audio: bool | None = None,
+        supports_video: bool | None = None,
     ):
         from nanobot.config.schema import ToolsConfig
 
@@ -222,6 +231,10 @@ class AgentLoop:
         self.tools_config = _tc
         self.web_config = _tc.web
         self.exec_config = _tc.exec
+        self.input_limits = input_limits or _tc.input_limits
+        self._supports_vision = supports_vision
+        self._supports_audio = supports_audio
+        self._supports_video = supports_video
         self._image_generation_provider_configs = dict(image_generation_provider_configs or {})
         if (
             image_generation_provider_config is not None
@@ -235,8 +248,13 @@ class AgentLoop:
         self._pending_turn_latency_ms: dict[str, int] = {}
         self._extra_hooks: list[AgentHook] = hooks or []
 
-        self.context = ContextBuilder(workspace, timezone=timezone, disabled_skills=disabled_skills)
+        self.context = ContextBuilder(workspace, timezone=timezone, disabled_skills=disabled_skills, input_limits=self.input_limits)
         self.sessions = session_manager or SessionManager(workspace)
+        self._webui_turns = WebuiTurnCoordinator(
+            bus=self.bus,
+            sessions=self.sessions,
+            schedule_background=lambda coro: self._schedule_background(coro),
+        )
         self.tools = ToolRegistry()
         # One file-read/write tracker per logical session. The tool registry is
         # shared by this loop, so tools resolve the active state via contextvars.
@@ -356,6 +374,10 @@ class AgentLoop:
             model_preset=defaults.model_preset,
             provider_snapshot_loader=provider_snapshot_loader,
             preset_snapshot_loader=preset_snapshot_loader,
+            input_limits=config.tools.input_limits,
+            supports_vision=defaults.supports_vision(defaults.model),
+            supports_audio=defaults.supports_audio(defaults.model),
+            supports_video=defaults.supports_video(defaults.model),
             **extra,
         )
 
@@ -524,34 +546,7 @@ class AgentLoop:
         self, msg: InboundMessage
     ) -> Callable[..., Awaitable[None]]:
         """Build a progress callback that publishes to the message bus."""
-
-        async def _bus_progress(
-            content: str,
-            *,
-            tool_hint: bool = False,
-            tool_events: list[dict[str, Any]] | None = None,
-            reasoning: bool = False,
-            reasoning_end: bool = False,
-        ) -> None:
-            meta = dict(msg.metadata or {})
-            meta["_progress"] = True
-            meta["_tool_hint"] = tool_hint
-            if reasoning:
-                meta["_reasoning_delta"] = True
-            if reasoning_end:
-                meta["_reasoning_end"] = True
-            if tool_events:
-                meta["_tool_events"] = tool_events
-            await self.bus.publish_outbound(
-                OutboundMessage(
-                    channel=msg.channel,
-                    chat_id=msg.chat_id,
-                    content=content,
-                    metadata=meta,
-                )
-            )
-
-        return _bus_progress
+        return build_bus_progress_callback(self.bus, msg)
 
     async def _build_retry_wait_callback(
         self, msg: InboundMessage
@@ -611,6 +606,9 @@ class AgentLoop:
             sender_id=msg.sender_id,
             session_summary=pending_summary,
             session_metadata=session.metadata,
+            supports_vision=self._supports_vision,
+            supports_audio=self._supports_audio,
+            supports_video=self._supports_video,
         )
 
     async def _dispatch_command_inline(
@@ -938,38 +936,12 @@ class AgentLoop:
                             content="", metadata=msg.metadata or {},
                         ))
                     if msg.channel == "websocket":
-                        # Signal that the turn is fully complete (all tools executed,
-                        # final text streamed).  This lets WS clients know when to
-                        # definitively stop the loading indicator.
                         turn_lat = self._pending_turn_latency_ms.pop(session_key, None)
-                        turn_metadata: dict[str, Any] = {**msg.metadata, "_turn_end": True}
-                        if turn_lat is not None:
-                            turn_metadata["latency_ms"] = int(turn_lat)
-                        sess_turn = self.sessions.get_or_create(session_key)
-                        turn_metadata["goal_state"] = goal_state_ws_blob(sess_turn.metadata)
-                        await self.bus.publish_outbound(OutboundMessage(
-                            channel=msg.channel, chat_id=msg.chat_id,
-                            content="", metadata=turn_metadata,
-                        ))
-                        if msg.metadata.get("webui") is True:
-                            async def _generate_title_and_notify() -> None:
-                                generated = await maybe_generate_webui_title_after_turn(
-                                    channel=msg.channel,
-                                    metadata=msg.metadata,
-                                    sessions=self.sessions,
-                                    session_key=session_key,
-                                    provider=self.provider,
-                                    model=self.model,
-                                )
-                                if generated:
-                                    await self.bus.publish_outbound(OutboundMessage(
-                                        channel=msg.channel,
-                                        chat_id=msg.chat_id,
-                                        content="",
-                                        metadata={**msg.metadata, "_session_updated": True},
-                                    ))
-
-                            self._schedule_background(_generate_title_and_notify())
+                        await self._webui_turns.handle_turn_end(
+                            msg,
+                            session_key=session_key,
+                            latency_ms=turn_lat,
+                        )
                 except asyncio.CancelledError:
                     logger.info("Task cancelled for session {}", session_key)
                     # Preserve partial context from the interrupted turn so
@@ -1021,8 +993,9 @@ class AgentLoop:
                         "Re-published {} leftover message(s) to bus for session {}",
                         leftover, session_key,
                     )
-            await publish_turn_run_status(self.bus, msg, "idle")
+            await self._webui_turns.publish_run_status(msg, "idle")
             self._pending_turn_latency_ms.pop(session_key, None)
+            self._webui_turns.discard(session_key)
 
     async def close_mcp(self) -> None:
         """Drain pending background archives, then close MCP connections."""
@@ -1101,6 +1074,9 @@ class AgentLoop:
             sender_id=msg.sender_id,
             session_summary=pending,
             session_metadata=session.metadata,
+            supports_vision=self._supports_vision,
+            supports_audio=self._supports_audio,
+            supports_video=self._supports_video,
         )
         t_wall = time.time()
         final_content, _, all_msgs, stop_reason, _ = await self._run_agent_loop(
@@ -1233,7 +1209,6 @@ class AgentLoop:
         all_msgs: list[dict[str, Any]],
         stop_reason: str,
         had_injections: bool,
-        generated_media: list[str],
         on_stream: Callable[[str], Awaitable[None]] | None,
         *,
         turn_latency_ms: int | None = None,
@@ -1257,7 +1232,6 @@ class AgentLoop:
             channel=msg.channel,
             chat_id=msg.chat_id,
             content=final_content,
-            media=generated_media,
             metadata=meta,
         )
 
@@ -1338,6 +1312,11 @@ class AgentLoop:
             "include_timestamps": True,
         }
         ctx.history = ctx.session.get_history(**_hist_kwargs)
+        self._webui_turns.capture_title_context(
+            ctx.session_key,
+            ctx.msg,
+            self.llm_runtime(),
+        )
 
         ctx.initial_messages = self._build_initial_messages(
             ctx.msg, ctx.session, ctx.history, ctx.pending_summary
@@ -1354,7 +1333,7 @@ class AgentLoop:
         return "ok"
 
     async def _state_run(self, ctx: TurnContext) -> str:
-        await publish_turn_run_status(self.bus, ctx.msg, "running")
+        await self._webui_turns.publish_run_status(ctx.msg, "running")
         result = await self._run_agent_loop(
             ctx.initial_messages,
             on_progress=ctx.on_progress,
@@ -1382,11 +1361,6 @@ class AgentLoop:
             ctx.final_content = EMPTY_FINAL_RESPONSE_MESSAGE
 
         ctx.save_skip = 1 + len(ctx.history) + (1 if ctx.user_persisted_early else 0)
-        skip_msgs = ctx.all_messages[ctx.save_skip:]
-        ctx.generated_media = generated_image_paths_from_messages(skip_msgs)
-        mt = self.tools.get("message")
-        extra = getattr(mt, "turn_delivered_media_paths", lambda: [])() if mt else []
-        merge_turn_media_into_last_assistant(ctx.all_messages, ctx.generated_media, extra)
 
         ctx.turn_latency_ms = max(0, int((time.time() - ctx.turn_wall_started_at) * 1000))
         self._save_turn(
@@ -1414,7 +1388,6 @@ class AgentLoop:
             ctx.all_messages,
             ctx.stop_reason,
             ctx.had_injections,
-            ctx.generated_media,
             ctx.on_stream,
             turn_latency_ms=ctx.turn_latency_ms,
         )
@@ -1449,6 +1422,10 @@ class AgentLoop:
                 filtered.append({"type": "text", "text": image_placeholder_text(path)})
                 continue
 
+            if block.get("type") in ("input_audio", "video_url"):
+                filtered.append(LLMProvider._media_placeholder(block["type"], block))
+                continue
+
             if block.get("type") == "text" and isinstance(block.get("text"), str):
                 text = block["text"]
                 if should_truncate_text and len(text) > self.max_tool_result_chars:

nanobot/agent/memory.py

@@ -678,11 +678,18 @@ class Consolidator:
         The budget reserves space for completion tokens and a safety buffer
         so the LLM request never exceeds the context window.
         """
-        if not session.messages or self.context_window_tokens <= 0:
+        if self.context_window_tokens <= 0:
             return
 
         lock = self.get_lock(session.key)
         async with lock:
+            # Refresh session reference: AutoCompact may have replaced it.
+            fresh = self.sessions.get_or_create(session.key)
+            if fresh is not session:
+                session = fresh
+            if not session.messages:
+                return
+
             budget = self._input_token_budget
             target = int(budget * self.consolidation_ratio)
             last_summary = await self._consolidate_replay_overflow(
@@ -769,6 +776,74 @@ class Consolidator:
             # the summary injection strategy with AutoCompact._archive().
             self._persist_last_summary(session, last_summary)
 
+    async def compact_idle_session(
+        self,
+        session_key: str,
+        max_suffix: int = 8,
+    ) -> str | None:
+        """Hard-truncate an idle session under the consolidation lock.
+
+        Used by AutoCompact so all session mutation goes through a single
+        lock-protected path.  Returns the summary text on success, ``None``
+        if the LLM failed (raw_archive fallback), or ``""`` if there was
+        nothing to archive.
+        """
+        lock = self.get_lock(session_key)
+        async with lock:
+            self.sessions.invalidate(session_key)
+            session = self.sessions.get_or_create(session_key)
+
+            tail = list(session.messages[session.last_consolidated:])
+            if not tail:
+                session.updated_at = datetime.now()
+                self.sessions.save(session)
+                return ""
+
+            probe = Session(
+                key=session.key,
+                messages=tail.copy(),
+                created_at=session.created_at,
+                updated_at=session.updated_at,
+                metadata={},
+                last_consolidated=0,
+            )
+            probe.retain_recent_legal_suffix(max_suffix)
+            kept = probe.messages
+            cut = len(tail) - len(kept)
+            archive_msgs = tail[:cut]
+
+            if not archive_msgs and not kept:
+                session.updated_at = datetime.now()
+                self.sessions.save(session)
+                return ""
+
+            last_active = session.updated_at
+            summary: str | None = ""
+            if archive_msgs:
+                summary = await self.archive(archive_msgs)
+
+            if summary and summary != "(nothing)":
+                session.metadata["_last_summary"] = {
+                    "text": summary,
+                    "last_active": last_active.isoformat(),
+                }
+
+            session.messages = kept
+            session.last_consolidated = 0
+            session.updated_at = datetime.now()
+            self.sessions.save(session)
+
+            if archive_msgs:
+                logger.info(
+                    "Idle-session compact for {}: archived={}, kept={}, summary={}",
+                    session_key,
+                    len(archive_msgs),
+                    len(kept),
+                    bool(summary),
+                )
+
+            return summary
+
 
 # ---------------------------------------------------------------------------
 # Dream — heavyweight cron-scheduled memory consolidation

nanobot/agent/runner.py

@@ -15,6 +15,13 @@ from loguru import logger
 from nanobot.agent.hook import AgentHook, AgentHookContext
 from nanobot.agent.tools.registry import ToolRegistry
 from nanobot.providers.base import LLMProvider, LLMResponse, ToolCallRequest
+from nanobot.utils.file_edit_events import (
+    build_file_edit_end_event,
+    build_file_edit_error_event,
+    build_file_edit_start_event,
+    prepare_file_edit_tracker,
+    StreamingFileEditTracker,
+)
 from nanobot.utils.helpers import (
     IncrementalThinkExtractor,
     build_assistant_message,
@@ -26,6 +33,10 @@ from nanobot.utils.helpers import (
     strip_think,
     truncate_text,
 )
+from nanobot.utils.progress_events import (
+    invoke_file_edit_progress,
+    on_progress_accepts_file_edit_events,
+)
 from nanobot.utils.prompt_templates import render_template
 from nanobot.utils.runtime import (
     EMPTY_FINAL_RESPONSE_MESSAGE,
@@ -619,6 +630,24 @@ class AgentRunner:
         )
 
         progress_state: dict[str, bool] | None = None
+        live_file_edits: StreamingFileEditTracker | None = None
+
+        if (
+            spec.progress_callback is not None
+            and on_progress_accepts_file_edit_events(spec.progress_callback)
+        ):
+            async def _emit_live_file_edits(events: list[dict[str, Any]]) -> None:
+                await invoke_file_edit_progress(spec.progress_callback, events)
+
+            live_file_edits = StreamingFileEditTracker(
+                workspace=spec.workspace,
+                tools=spec.tools,
+                emit=_emit_live_file_edits,
+            )
+
+        async def _tool_call_delta(delta: dict[str, Any]) -> None:
+            if live_file_edits is not None:
+                await live_file_edits.update(delta)
 
         if wants_streaming:
             async def _stream(delta: str) -> None:
@@ -636,6 +665,7 @@ class AgentRunner:
                 **kwargs,
                 on_content_delta=_stream,
                 on_thinking_delta=_thinking,
+                on_tool_call_delta=_tool_call_delta if live_file_edits is not None else None,
             )
         elif wants_progress_streaming:
             stream_buf = ""
@@ -665,6 +695,7 @@ class AgentRunner:
             coro = self.provider.chat_stream_with_retry(
                 **kwargs,
                 on_content_delta=_stream_progress,
+                on_tool_call_delta=_tool_call_delta if live_file_edits is not None else None,
             )
         else:
             coro = self.provider.chat_with_retry(**kwargs)
@@ -679,6 +710,14 @@ class AgentRunner:
                 await coro if outer_timeout_s is None
                 else await asyncio.wait_for(coro, timeout=outer_timeout_s)
             )
+            if live_file_edits is not None:
+                await live_file_edits.flush()
+                if response.should_execute_tools:
+                    live_file_edits.apply_final_call_ids(response.tool_calls)
+                await live_file_edits.error_unmatched(
+                    response.tool_calls if response.should_execute_tools else [],
+                    "Tool call did not complete.",
+                )
         except asyncio.TimeoutError:
             if outer_timeout_s is None:
                 return LLMResponse(
@@ -813,6 +852,30 @@ class AgentRunner:
             return prep_error + hint, event, (
                 RuntimeError(prep_error) if spec.fail_on_tool_error else None
             )
+        emit_file_edit_events = (
+            spec.progress_callback is not None
+            and on_progress_accepts_file_edit_events(spec.progress_callback)
+        )
+        progress_callback = spec.progress_callback if emit_file_edit_events else None
+        file_edit_tracker = (
+            prepare_file_edit_tracker(
+                call_id=tool_call.id,
+                tool_name=tool_call.name,
+                tool=tool,
+                workspace=spec.workspace,
+                params=params if isinstance(params, dict) else None,
+            )
+            if progress_callback is not None
+            else None
+        )
+        if file_edit_tracker is not None and progress_callback is not None:
+            await invoke_file_edit_progress(
+                progress_callback,
+                [build_file_edit_start_event(
+                    file_edit_tracker,
+                    params if isinstance(params, dict) else None,
+                )],
+            )
         try:
             if tool is not None:
                 result = await tool.execute(**params)
@@ -821,6 +884,11 @@ class AgentRunner:
         except asyncio.CancelledError:
             raise
         except BaseException as exc:
+            if file_edit_tracker is not None and progress_callback is not None:
+                await invoke_file_edit_progress(
+                    progress_callback,
+                    [build_file_edit_error_event(file_edit_tracker, str(exc))],
+                )
             event = {
                 "name": tool_call.name,
                 "status": "error",
@@ -842,6 +910,11 @@ class AgentRunner:
             return payload, event, None
 
         if isinstance(result, str) and result.startswith("Error"):
+            if file_edit_tracker is not None and progress_callback is not None:
+                await invoke_file_edit_progress(
+                    progress_callback,
+                    [build_file_edit_error_event(file_edit_tracker, result)],
+                )
             event = {
                 "name": tool_call.name,
                 "status": "error",
@@ -860,6 +933,15 @@ class AgentRunner:
                 return result + hint, event, RuntimeError(result)
             return result + hint, event, None
 
+        if file_edit_tracker is not None and progress_callback is not None:
+            await invoke_file_edit_progress(
+                progress_callback,
+                [build_file_edit_end_event(
+                    file_edit_tracker,
+                    params if isinstance(params, dict) else None,
+                )],
+            )
+
         detail = "" if result is None else str(result)
         detail = detail.replace("\n", " ").strip()
         if not detail:

nanobot/agent/tools/image_generation.py

@@ -17,9 +17,9 @@ from nanobot.agent.tools.schema import (
 from nanobot.config.paths import get_media_dir
 from nanobot.config.schema import Base
 from nanobot.providers.image_generation import (
-    AIHubMixImageGenerationClient,
     ImageGenerationError,
-    OpenRouterImageGenerationClient,
+    ImageGenerationProvider,
+    get_image_gen_provider,
 )
 from nanobot.utils.artifacts import (
     ArtifactError,
@@ -117,27 +117,24 @@ class ImageGenerationTool(Tool):
     def _provider_config(self) -> ProviderConfig | None:
         return self.provider_configs.get(self.config.provider)
 
-    def _provider_client(self) -> OpenRouterImageGenerationClient | AIHubMixImageGenerationClient | None:
+    def _provider_client(self) -> ImageGenerationProvider | None:
         provider = self._provider_config()
+        cls = get_image_gen_provider(self.config.provider)
+        if cls is None:
+            return None
         kwargs = {
             "api_key": provider.api_key if provider else None,
             "api_base": provider.api_base if provider else None,
             "extra_headers": provider.extra_headers if provider else None,
             "extra_body": provider.extra_body if provider else None,
         }
-        if self.config.provider == "openrouter":
-            return OpenRouterImageGenerationClient(**kwargs)
-        if self.config.provider == "aihubmix":
-            return AIHubMixImageGenerationClient(**kwargs)
-        return None
+        return cls(**kwargs)
 
     def _missing_api_key_error(self) -> str:
-        provider = self.config.provider
-        if provider == "openrouter":
-            return "Error: OpenRouter API key is not configured. Set providers.openrouter.apiKey."
-        if provider == "aihubmix":
-            return "Error: AIHubMix API key is not configured. Set providers.aihubmix.apiKey."
-        return f"Error: {provider} API key is not configured."
+        cls = get_image_gen_provider(self.config.provider)
+        if cls and cls.missing_key_message:
+            return f"Error: {cls.missing_key_message}"
+        return f"Error: {self.config.provider} API key is not configured."
 
     def _resolve_reference_image(self, value: str) -> str:
         raw_path = Path(value).expanduser()

nanobot/agent/tools/message.py

@@ -31,8 +31,8 @@ from nanobot.config.paths import get_workspace_path
         media=ArraySchema(
             StringSchema(""),
             description=(
-                "Optional list of existing file paths to attach for proactive or cross-channel delivery. "
-                "Do not use this to resend generate_image outputs in the current chat."
+                "Optional list of existing file paths to attach. "
+                "Use artifact paths returned by generate_image here when delivering generated images."
             ),
         ),
         buttons=ArraySchema(
@@ -140,8 +140,8 @@ class MessageTool(Tool, ContextAware):
             "Do not use this for the normal reply in the current chat: answer naturally instead. "
             "If channel/chat_id would target the current runtime conversation, do not call this tool "
             "unless the user explicitly asked you to proactively send an existing file attachment. "
-            "When generate_image creates images in the current chat, the final assistant reply "
-            "automatically attaches them; do not call message just to announce or resend them. "
+            "When generate_image creates images in the current chat, use the message tool "
+            "with the artifact paths in the media parameter to deliver the images to the user. "
             "For proactive attachment delivery, use the 'media' parameter with file paths. "
             "Do NOT use read_file to send files — that only reads content for your own analysis."
         )

nanobot/channels/feishu.py

@@ -172,19 +172,22 @@ def _extract_element_content(element: dict) -> list[str]:
     return parts
 
 
-def _extract_post_content(content_json: dict) -> tuple[str, list[str]]:
-    """Extract text and image keys from Feishu post (rich text) message.
+def _extract_post_content(content_json: dict) -> tuple[str, list[str], list[dict]]:
+    """Extract text and media info from Feishu post (rich text) message.
 
     Handles three payload shapes:
     - Direct:    {"title": "...", "content": [[...]]}
     - Localized: {"zh_cn": {"title": "...", "content": [...]}}
     - Wrapped:   {"post": {"zh_cn": {"title": "...", "content": [...]}}}
+
+    Returns (text, image_keys, media_items) where media_items is a list of
+    {"tag": "media", "file_key": "..."} dicts for video/file attachments.
     """
 
-    def _parse_block(block: dict) -> tuple[str | None, list[str]]:
+    def _parse_block(block: dict) -> tuple[str | None, list[str], list[dict]]:
         if not isinstance(block, dict) or not isinstance(block.get("content"), list):
-            return None, []
-        texts, images = [], []
+            return None, [], []
+        texts, images, medias = [], [], []
         if title := block.get("title"):
             texts.append(title)
         for row in block["content"]:
@@ -204,43 +207,36 @@ def _extract_post_content(content_json: dict) -> tuple[str, list[str]]:
                     texts.append(f"\n```{lang}\n{code_text}\n```\n")
                 elif tag == "img" and (key := el.get("image_key")):
                     images.append(key)
-        return (" ".join(texts).strip() or None), images
+                elif tag == "media" and el.get("file_key"):
+                    medias.append({"tag": "media", "file_key": el["file_key"]})
+        return (" ".join(texts).strip() or None), images, medias
 
     # Unwrap optional {"post": ...} envelope
     root = content_json
     if isinstance(root, dict) and isinstance(root.get("post"), dict):
         root = root["post"]
     if not isinstance(root, dict):
-        return "", []
+        return "", [], []
 
     # Direct format
     if "content" in root:
-        text, imgs = _parse_block(root)
-        if text or imgs:
-            return text or "", imgs
+        text, imgs, medias = _parse_block(root)
+        if text or imgs or medias:
+            return text or "", imgs, medias
 
     # Localized: prefer known locales, then fall back to any dict child
     for key in ("zh_cn", "en_us", "ja_jp"):
         if key in root:
-            text, imgs = _parse_block(root[key])
-            if text or imgs:
-                return text or "", imgs
+            text, imgs, medias = _parse_block(root[key])
+            if text or imgs or medias:
+                return text or "", imgs, medias
     for val in root.values():
         if isinstance(val, dict):
-            text, imgs = _parse_block(val)
-            if text or imgs:
-                return text or "", imgs
+            text, imgs, medias = _parse_block(val)
+            if text or imgs or medias:
+                return text or "", imgs, medias
 
-    return "", []
-
-
-def _extract_post_text(content_json: dict) -> str:
-    """Extract plain text from Feishu post (rich text) message content.
-
-    Legacy wrapper for _extract_post_content, returns only text.
-    """
-    text, _ = _extract_post_content(content_json)
-    return text
+    return "", [], []
 
 
 class FeishuConfig(Base):
@@ -1156,7 +1152,7 @@ class FeishuChannel(BaseChannel):
             if msg_type == "text":
                 text = content_json.get("text", "").strip()
             elif msg_type == "post":
-                text, _ = _extract_post_content(content_json)
+                text, _, _ = _extract_post_content(content_json)
                 text = text.strip()
             else:
                 text = ""
@@ -1751,7 +1747,7 @@ class FeishuChannel(BaseChannel):
                     content_parts.append(text)
 
             elif msg_type == "post":
-                text, image_keys = _extract_post_content(content_json)
+                text, image_keys, media_items = _extract_post_content(content_json)
                 if text:
                     content_parts.append(text)
                 # Download images embedded in post
@@ -1762,6 +1758,14 @@ class FeishuChannel(BaseChannel):
                     if file_path:
                         media_paths.append(file_path)
                     content_parts.append(content_text)
+                # Download media (video/file) embedded in post
+                for media_item in media_items:
+                    file_path, content_text = await self._download_and_save_media(
+                        "media", media_item, message_id
+                    )
+                    if file_path:
+                        media_paths.append(file_path)
+                    content_parts.append(content_text)
 
             elif msg_type in ("image", "audio", "file", "media"):
                 file_path, content_text = await self._download_and_save_media(

nanobot/channels/websocket.py

@@ -37,15 +37,27 @@ from nanobot.command.builtin import builtin_command_palette
 from nanobot.config.paths import get_media_dir
 from nanobot.config.schema import Base
 from nanobot.session.goal_state import goal_state_ws_blob
+from nanobot.session.webui_turns import websocket_turn_wall_started_at
 from nanobot.utils.helpers import safe_filename
 from nanobot.utils.media_decode import (
     FileSizeExceeded,
     save_base64_data_url,
 )
 from nanobot.utils.subagent_channel_display import scrub_subagent_messages_for_channel
-from nanobot.utils.webui_thread_disk import delete_webui_thread
-from nanobot.utils.webui_transcript import append_transcript_object, build_webui_thread_response
-from nanobot.utils.webui_turn_helpers import websocket_turn_wall_started_at
+from nanobot.webui.settings_api import (
+    WebUISettingsError,
+    settings_payload,
+    update_agent_settings,
+    update_image_generation_settings,
+    update_provider_settings,
+    update_web_search_settings,
+)
+from nanobot.webui.sidebar_state import (
+    read_webui_sidebar_state,
+    write_webui_sidebar_state,
+)
+from nanobot.webui.thread_disk import delete_webui_thread
+from nanobot.webui.transcript import append_transcript_object, build_webui_thread_response
 
 if TYPE_CHECKING:
     from nanobot.session.manager import SessionManager
@@ -222,28 +234,6 @@ def _query_first(query: dict[str, list[str]], key: str) -> str | None:
     return values[0] if values else None
 
 
-def _mask_secret_hint(secret: str | None) -> str | None:
-    if not secret:
-        return None
-    if len(secret) <= 8:
-        return "••••"
-    return f"{secret[:4]}••••{secret[-4:]}"
-
-
-_WEB_SEARCH_PROVIDER_OPTIONS: tuple[dict[str, str], ...] = (
-    {"name": "duckduckgo", "label": "DuckDuckGo", "credential": "none"},
-    {"name": "brave", "label": "Brave Search", "credential": "api_key"},
-    {"name": "tavily", "label": "Tavily", "credential": "api_key"},
-    {"name": "searxng", "label": "SearXNG", "credential": "base_url"},
-    {"name": "jina", "label": "Jina", "credential": "api_key"},
-    {"name": "kagi", "label": "Kagi", "credential": "api_key"},
-    {"name": "olostep", "label": "Olostep", "credential": "api_key"},
-)
-_WEB_SEARCH_PROVIDER_BY_NAME = {
-    provider["name"]: provider for provider in _WEB_SEARCH_PROVIDER_OPTIONS
-}
-
-
 def _parse_inbound_payload(raw: str) -> str | None:
     """Parse a client frame into text; return None for empty or unrecognized content."""
     text = raw.strip()
@@ -482,6 +472,7 @@ class WebSocketChannel(BaseChannel):
             static_dist_path.resolve() if static_dist_path is not None else None
         )
         self._runtime_model_name = runtime_model_name
+        self._settings_restart_sections: set[str] = set()
         # Process-local secret used to HMAC-sign media URLs. The signed URL is
         # the capability — anyone who holds a valid URL can fetch that one
         # file, nothing else. The secret regenerates on restart so links
@@ -644,6 +635,12 @@ class WebSocketChannel(BaseChannel):
         if got == "/api/commands":
             return self._handle_commands(request)
 
+        if got == "/api/webui/sidebar-state":
+            return self._handle_webui_sidebar_state(request)
+
+        if got == "/api/webui/sidebar-state/update":
+            return self._handle_webui_sidebar_state_update(request)
+
         if got == "/api/settings/update":
             return self._handle_settings_update(request)
 
@@ -653,6 +650,9 @@ class WebSocketChannel(BaseChannel):
         if got == "/api/settings/web-search/update":
             return self._handle_settings_web_search_update(request)
 
+        if got == "/api/settings/image-generation/update":
+            return self._handle_settings_image_generation_update(request)
+
         m = re.match(r"^/api/sessions/([^/]+)/messages$", got)
         if m:
             return self._handle_session_messages(request, m.group(1))
@@ -764,215 +764,115 @@ class WebSocketChannel(BaseChannel):
         sessions = self._session_manager.list_sessions()
         # Sidebar/chat listing for WS-backed sessions only — CLI / Slack / etc.
         # keys are not intended for resume over this HTTP surface.
-        cleaned = [
-            {k: v for k, v in s.items() if k != "path"}
-            for s in sessions
-            if isinstance(s.get("key"), str) and s["key"].startswith("websocket:")
-        ]
-        return _http_json_response({"sessions": cleaned})
-
-    def _settings_payload(self, *, requires_restart: bool = False) -> dict[str, Any]:
-        from nanobot.config.loader import get_config_path, load_config
-        from nanobot.providers.registry import PROVIDERS, find_by_name
-
-        config = load_config()
-        defaults = config.agents.defaults
-        provider_name = config.get_provider_name(defaults.model) or defaults.provider
-        provider = config.get_provider(defaults.model)
-        selected_provider = provider_name
-        if defaults.provider != "auto":
-            spec = find_by_name(defaults.provider)
-            selected_provider = spec.name if spec else provider_name
-        providers = []
-        for spec in PROVIDERS:
-            provider_config = getattr(config.providers, spec.name, None)
-            if provider_config is None or spec.is_oauth or spec.is_local:
+        cleaned = []
+        for s in sessions:
+            key = s.get("key")
+            if not (isinstance(key, str) and key.startswith("websocket:")):
                 continue
-            providers.append(
-                {
-                    "name": spec.name,
-                    "label": spec.label,
-                    "configured": bool(provider_config.api_key),
-                    "api_key_hint": _mask_secret_hint(provider_config.api_key),
-                    "api_base": provider_config.api_base,
-                    "default_api_base": spec.default_api_base or None,
-                }
-            )
-        search_config = config.tools.web.search
-        search_provider = (
-            search_config.provider
-            if search_config.provider in _WEB_SEARCH_PROVIDER_BY_NAME
-            else "duckduckgo"
-        )
-        return {
-            "agent": {
-                "model": defaults.model,
-                "provider": selected_provider,
-                "resolved_provider": provider_name,
-                "has_api_key": bool(provider and provider.api_key),
-            },
-            "providers": providers,
-            "web_search": {
-                "provider": search_provider,
-                "api_key_hint": _mask_secret_hint(search_config.api_key),
-                "base_url": search_config.base_url or None,
-                "providers": list(_WEB_SEARCH_PROVIDER_OPTIONS),
-            },
-            "runtime": {
-                "config_path": str(get_config_path().expanduser()),
-            },
-            "requires_restart": requires_restart,
-        }
+            row = {k: v for k, v in s.items() if k != "path"}
+            chat_id = key.split(":", 1)[1]
+            started_at = websocket_turn_wall_started_at(chat_id)
+            if started_at is not None:
+                row["run_started_at"] = started_at
+            cleaned.append(row)
+        return _http_json_response({"sessions": cleaned})
 
     def _handle_settings(self, request: WsRequest) -> Response:
         if not self._check_api_token(request):
             return _http_error(401, "Unauthorized")
-        return _http_json_response(self._settings_payload())
+        return _http_json_response(self._with_settings_restart_state(settings_payload()))
+
+    def _with_settings_restart_state(
+        self,
+        payload: dict[str, Any],
+        *,
+        section: str | None = None,
+    ) -> dict[str, Any]:
+        """Keep restart-required state alive for this gateway process."""
+        if section and payload.get("requires_restart"):
+            self._settings_restart_sections.add(section)
+        if self._settings_restart_sections:
+            payload = dict(payload)
+            payload["requires_restart"] = True
+            payload["restart_required_sections"] = sorted(self._settings_restart_sections)
+        else:
+            payload = dict(payload)
+            payload["restart_required_sections"] = []
+        return payload
 
     def _handle_commands(self, request: WsRequest) -> Response:
         if not self._check_api_token(request):
             return _http_error(401, "Unauthorized")
         return _http_json_response({"commands": builtin_command_palette()})
 
+    def _handle_webui_sidebar_state(self, request: WsRequest) -> Response:
+        if not self._check_api_token(request):
+            return _http_error(401, "Unauthorized")
+        return _http_json_response(read_webui_sidebar_state())
+
+    def _handle_webui_sidebar_state_update(self, request: WsRequest) -> Response:
+        if not self._check_api_token(request):
+            return _http_error(401, "Unauthorized")
+        query = _parse_query(request.path)
+        raw_state = _query_first(query, "state")
+        if raw_state is None:
+            return _http_error(400, "missing state")
+        try:
+            decoded = json.loads(raw_state)
+        except json.JSONDecodeError:
+            return _http_error(400, "state must be JSON")
+        if not isinstance(decoded, dict):
+            return _http_error(400, "state must be an object")
+        try:
+            state = write_webui_sidebar_state(decoded)
+        except ValueError as e:
+            return _http_error(400, str(e))
+        except OSError:
+            self.logger.exception("failed to write webui sidebar state")
+            return _http_error(500, "failed to write sidebar state")
+        return _http_json_response(state)
+
     def _handle_settings_update(self, request: WsRequest) -> Response:
         if not self._check_api_token(request):
             return _http_error(401, "Unauthorized")
-        from nanobot.config.loader import load_config, save_config
-        from nanobot.providers.registry import find_by_name
-
         query = _parse_query(request.path)
-        config = load_config()
-        defaults = config.agents.defaults
-        changed = False
-
-        model = _query_first(query, "model")
-        if model is not None:
-            model = model.strip()
-            if not model:
-                return _http_error(400, "model is required")
-            if defaults.model != model:
-                defaults.model = model
-                changed = True
-
-        provider = _query_first(query, "provider")
-        if provider is not None:
-            provider = provider.strip()
-            if not provider:
-                return _http_error(400, "provider is required")
-            if find_by_name(provider) is None:
-                return _http_error(400, "unknown provider")
-            provider_config = getattr(config.providers, provider, None)
-            if provider_config is None or not provider_config.api_key:
-                return _http_error(400, "provider is not configured")
-            if defaults.provider != provider:
-                defaults.provider = provider
-                changed = True
-
-        if changed:
-            save_config(config)
-        # LLM provider/model changes are hot-reloaded by AgentLoop before each
-        # new turn via the provider snapshot loader, so a restart is unnecessary.
-        return _http_json_response(self._settings_payload(requires_restart=False))
+        try:
+            payload = update_agent_settings(query)
+        except WebUISettingsError as e:
+            return _http_error(e.status, e.message)
+        return _http_json_response(
+            self._with_settings_restart_state(payload, section="runtime")
+        )
 
     def _handle_settings_provider_update(self, request: WsRequest) -> Response:
         if not self._check_api_token(request):
             return _http_error(401, "Unauthorized")
-        from nanobot.config.loader import load_config, save_config
-        from nanobot.providers.registry import find_by_name
-
         query = _parse_query(request.path)
-        provider_name = (_query_first(query, "provider") or "").strip()
-        if not provider_name:
-            return _http_error(400, "provider is required")
-        spec = find_by_name(provider_name)
-        if spec is None or spec.is_oauth or spec.is_local:
-            return _http_error(400, "unknown provider")
-
-        config = load_config()
-        provider_config = getattr(config.providers, spec.name, None)
-        if provider_config is None:
-            return _http_error(400, "unknown provider")
-
-        changed = False
-        if "api_key" in query or "apiKey" in query:
-            api_key = _query_first(query, "api_key")
-            if api_key is None:
-                api_key = _query_first(query, "apiKey")
-            api_key = (api_key or "").strip() or None
-            if provider_config.api_key != api_key:
-                provider_config.api_key = api_key
-                changed = True
-
-        if "api_base" in query or "apiBase" in query:
-            api_base = _query_first(query, "api_base")
-            if api_base is None:
-                api_base = _query_first(query, "apiBase")
-            api_base = (api_base or "").strip() or None
-            if provider_config.api_base != api_base:
-                provider_config.api_base = api_base
-                changed = True
-
-        if changed:
-            save_config(config)
-        # API key/base changes are picked up by the next provider snapshot refresh.
-        return _http_json_response(self._settings_payload(requires_restart=False))
+        try:
+            payload = update_provider_settings(query)
+        except WebUISettingsError as e:
+            return _http_error(e.status, e.message)
+        return _http_json_response(self._with_settings_restart_state(payload, section="image"))
 
     def _handle_settings_web_search_update(self, request: WsRequest) -> Response:
         if not self._check_api_token(request):
             return _http_error(401, "Unauthorized")
-        from nanobot.config.loader import load_config, save_config
-
         query = _parse_query(request.path)
-        provider_name = (_query_first(query, "provider") or "").strip().lower()
-        provider_option = _WEB_SEARCH_PROVIDER_BY_NAME.get(provider_name)
-        if provider_option is None:
-            return _http_error(400, "unknown web search provider")
+        try:
+            payload = update_web_search_settings(query)
+        except WebUISettingsError as e:
+            return _http_error(e.status, e.message)
+        return _http_json_response(self._with_settings_restart_state(payload, section="web"))
 
-        config = load_config()
-        search_config = config.tools.web.search
-        previous_provider = search_config.provider
-        changed = False
-
-        def set_value(attr: str, value: str | None) -> None:
-            nonlocal changed
-            if getattr(search_config, attr) != value:
-                setattr(search_config, attr, value)
-                changed = True
-
-        if search_config.provider != provider_name:
-            search_config.provider = provider_name
-            changed = True
-
-        credential = provider_option["credential"]
-        if credential == "none":
-            set_value("api_key", "")
-            set_value("base_url", "")
-        elif credential == "base_url":
-            base_url = _query_first(query, "base_url")
-            if base_url is None:
-                base_url = _query_first(query, "baseUrl")
-            base_url = base_url.strip() if base_url is not None else None
-            if not base_url and previous_provider == provider_name and search_config.base_url:
-                base_url = search_config.base_url
-            if not base_url:
-                return _http_error(400, "base_url is required")
-            set_value("base_url", base_url)
-            set_value("api_key", "")
-        else:
-            api_key = _query_first(query, "api_key")
-            if api_key is None:
-                api_key = _query_first(query, "apiKey")
-            api_key = api_key.strip() if api_key is not None else None
-            if not api_key and previous_provider == provider_name and search_config.api_key:
-                api_key = search_config.api_key
-            if not api_key:
-                return _http_error(400, "api_key is required")
-            set_value("api_key", api_key)
-            set_value("base_url", "")
-
-        if changed:
-            save_config(config)
-        return _http_json_response(self._settings_payload(requires_restart=False))
+    def _handle_settings_image_generation_update(self, request: WsRequest) -> Response:
+        if not self._check_api_token(request):
+            return _http_error(401, "Unauthorized")
+        query = _parse_query(request.path)
+        try:
+            payload = update_image_generation_settings(query)
+        except WebUISettingsError as e:
+            return _http_error(e.status, e.message)
+        return _http_json_response(self._with_settings_restart_state(payload, section="image"))
 
     @staticmethod
     def _is_websocket_channel_session_key(key: str) -> bool:
@@ -1581,6 +1481,7 @@ class WebSocketChannel(BaseChannel):
         if not conns:
             if (
                 msg.metadata.get("_progress")
+                or msg.metadata.get("_file_edit_events")
                 or msg.metadata.get("_turn_end")
                 or msg.metadata.get("_session_updated")
                 or msg.metadata.get("_goal_status")
@@ -1613,7 +1514,22 @@ class WebSocketChannel(BaseChannel):
             await self.send_turn_end(msg.chat_id, latency_ms=lat_i, goal_state=gs_blob)
             return
         if msg.metadata.get("_session_updated"):
-            await self.send_session_updated(msg.chat_id)
+            scope = msg.metadata.get("_session_update_scope")
+            await self.send_session_updated(
+                msg.chat_id,
+                scope=scope if isinstance(scope, str) else None,
+            )
+            return
+        if msg.metadata.get("_file_edit_events"):
+            payload: dict[str, Any] = {
+                "event": "file_edit",
+                "chat_id": msg.chat_id,
+                "edits": msg.metadata["_file_edit_events"],
+            }
+            self._try_append_webui_transcript(msg.chat_id, payload)
+            raw = json.dumps(payload, ensure_ascii=False)
+            for connection in conns:
+                await self._safe_send_to(connection, raw, label=" ")
             return
         text = msg.content
         payload: dict[str, Any] = {
@@ -1780,12 +1696,14 @@ class WebSocketChannel(BaseChannel):
         for connection in conns:
             await self._safe_send_to(connection, raw, label=" goal_status ")
 
-    async def send_session_updated(self, chat_id: str) -> None:
+    async def send_session_updated(self, chat_id: str, *, scope: str | None = None) -> None:
         """Notify clients that session metadata changed outside the main turn."""
         conns = list(self._subs.get(chat_id, ()))
         if not conns:
             return
         body: dict[str, Any] = {"event": "session_updated", "chat_id": chat_id}
+        if scope:
+            body["scope"] = scope
         raw = json.dumps(body, ensure_ascii=False)
         for connection in conns:
             await self._safe_send_to(connection, raw, label=" session_updated ")

nanobot/cli/commands.py

@@ -91,6 +91,8 @@ app = typer.Typer(
 
 console = Console()
 EXIT_COMMANDS = {"exit", "quit", "/exit", "/quit", ":q"}
+_REASONING_SENTENCE_ENDINGS = (".", "!", "?", "。", "！", "？")
+_REASONING_FLUSH_CHARS = 60
 
 # ---------------------------------------------------------------------------
 # CLI input: prompt_toolkit for editing, paste, history, and display
@@ -242,6 +244,35 @@ def _print_cli_progress_line(text: str, thinking: ThinkingSpinner | None, render
         target.print(f"  [dim]↳ {text}[/dim]")
 
 
+class _ReasoningBuffer:
+    def __init__(self) -> None:
+        self._text = ""
+
+    def add(self, text: str) -> str | None:
+        if not text:
+            return None
+        self._text += text
+        if self._should_flush(text):
+            return self.flush()
+        return None
+
+    def flush(self) -> str | None:
+        text = self._text.strip()
+        self._text = ""
+        return text or None
+
+    def clear(self) -> None:
+        self._text = ""
+
+    def _should_flush(self, text: str) -> bool:
+        stripped = text.rstrip()
+        return (
+            "\n" in text
+            or stripped.endswith(_REASONING_SENTENCE_ENDINGS)
+            or len(self._text) >= _REASONING_FLUSH_CHARS
+        )
+
+
 def _print_cli_reasoning(text: str, thinking: ThinkingSpinner | None, renderer: StreamRenderer | None = None) -> None:
     """Print reasoning/thinking content in a distinct style."""
     if not text.strip():
@@ -254,6 +285,16 @@ def _print_cli_reasoning(text: str, thinking: ThinkingSpinner | None, renderer:
         target.print(f"[dim italic]✻ {text}[/dim italic]")
 
 
+def _flush_cli_reasoning(
+    reasoning_buffer: _ReasoningBuffer,
+    thinking: ThinkingSpinner | None,
+    renderer: StreamRenderer | None = None,
+) -> None:
+    text = reasoning_buffer.flush()
+    if text:
+        _print_cli_reasoning(text, thinking, renderer)
+
+
 async def _print_interactive_progress_line(text: str, thinking: ThinkingSpinner | None, renderer: StreamRenderer | None = None) -> None:
     """Print an interactive progress line, pausing the spinner if needed."""
     if not text.strip():
@@ -272,6 +313,7 @@ async def _maybe_print_interactive_progress(
     thinking: ThinkingSpinner | None,
     channels_config: Any,
     renderer: StreamRenderer | None = None,
+    reasoning_buffer: _ReasoningBuffer | None = None,
 ) -> bool:
     metadata = msg.metadata or {}
     if metadata.get("_retry_wait"):
@@ -281,12 +323,24 @@ async def _maybe_print_interactive_progress(
     if not metadata.get("_progress"):
         return False
 
+    reasoning_buffer = reasoning_buffer or _ReasoningBuffer()
+
+    if metadata.get("_reasoning_end"):
+        if channels_config and not channels_config.show_reasoning:
+            reasoning_buffer.clear()
+        else:
+            _flush_cli_reasoning(reasoning_buffer, thinking, renderer)
+        return True
+
     is_tool_hint = metadata.get("_tool_hint", False)
     is_reasoning = metadata.get("_reasoning", False) or metadata.get("_reasoning_delta", False)
     if is_reasoning:
         if channels_config and not channels_config.show_reasoning:
+            reasoning_buffer.clear()
             return True
-        _print_cli_reasoning(msg.content, thinking, renderer)
+        text = reasoning_buffer.add(msg.content)
+        if text:
+            _print_cli_reasoning(text, thinking, renderer)
         return True
     if channels_config and is_tool_hint and not channels_config.send_tool_hints:
         return True
@@ -566,6 +620,7 @@ def serve(
 
     from nanobot.api.server import create_app
     from nanobot.bus.queue import MessageBus
+    from nanobot.providers.image_generation import image_gen_provider_configs
     from nanobot.session.manager import SessionManager
 
     if verbose:
@@ -585,10 +640,7 @@ def serve(
         agent_loop = AgentLoop.from_config(
             runtime_config, bus,
             session_manager=session_manager,
-            image_generation_provider_configs={
-                "openrouter": runtime_config.providers.openrouter,
-                "aihubmix": runtime_config.providers.aihubmix,
-            },
+            image_generation_provider_configs=image_gen_provider_configs(runtime_config),
         )
     except ValueError as exc:
         console.print(f"[red]Error: {exc}[/red]")
@@ -668,6 +720,7 @@ def _run_gateway(
     from nanobot.cron.types import CronJob
     from nanobot.heartbeat.service import HeartbeatService
     from nanobot.providers.factory import build_provider_snapshot, load_provider_snapshot
+    from nanobot.providers.image_generation import image_gen_provider_configs
     from nanobot.session.manager import SessionManager
 
     port = port if port is not None else config.gateway.port
@@ -698,10 +751,7 @@ def _run_gateway(
         context_window_tokens=provider_snapshot.context_window_tokens,
         cron_service=cron,
         session_manager=session_manager,
-        image_generation_provider_configs={
-            "openrouter": config.providers.openrouter,
-            "aihubmix": config.providers.aihubmix,
-        },
+        image_generation_provider_configs=image_gen_provider_configs(config),
         provider_snapshot_loader=load_provider_snapshot,
         runtime_model_publisher=lambda model, preset: publish_runtime_model_update(
             bus,
@@ -914,8 +964,7 @@ def _run_gateway(
     hb_cfg = config.gateway.heartbeat
     heartbeat = HeartbeatService(
         workspace=config.workspace_path,
-        provider=agent.provider,
-        model=agent.model,
+        llm_runtime=agent.llm_runtime,
         on_execute=on_heartbeat_execute,
         on_notify=on_heartbeat_notify,
         interval_s=hb_cfg.interval_s,
@@ -1069,6 +1118,7 @@ def agent(
 
     from nanobot.bus.queue import MessageBus
     from nanobot.cron.service import CronService
+    from nanobot.providers.image_generation import image_gen_provider_configs
 
     config = _load_runtime_config(config, workspace)
     sync_workspace_templates(config.workspace_path)
@@ -1092,6 +1142,7 @@ def agent(
         agent_loop = AgentLoop.from_config(
             config, bus,
             cron_service=cron,
+            image_generation_provider_configs=image_gen_provider_configs(config),
         )
     except ValueError as exc:
         console.print(f"[red]Error: {exc}[/red]")
@@ -1107,12 +1158,25 @@ def agent(
     _thinking: ThinkingSpinner | None = None
 
     def _make_progress(renderer: StreamRenderer | None = None):
+        reasoning_buffer = _ReasoningBuffer()
+
         async def _cli_progress(content: str, *, tool_hint: bool = False, reasoning: bool = False, **_kwargs: Any) -> None:
             ch = agent_loop.channels_config
+
+            if _kwargs.get("reasoning_end"):
+                if ch and not ch.show_reasoning:
+                    reasoning_buffer.clear()
+                else:
+                    _flush_cli_reasoning(reasoning_buffer, _thinking, renderer)
+                return
+
             if reasoning:
                 if ch and not ch.show_reasoning:
+                    reasoning_buffer.clear()
                     return
-                _print_cli_reasoning(content, _thinking, renderer)
+                text = reasoning_buffer.add(content)
+                if text:
+                    _print_cli_reasoning(text, _thinking, renderer)
                 return
             if ch and tool_hint and not ch.send_tool_hints:
                 return
@@ -1183,6 +1247,7 @@ def agent(
             turn_done.set()
             turn_response: list[tuple[str, dict]] = []
             renderer: StreamRenderer | None = None
+            reasoning_buffer = _ReasoningBuffer()
 
             async def _consume_outbound():
                 while True:
@@ -1208,6 +1273,7 @@ def agent(
                             renderer,
                             agent_loop.channels_config,
                             renderer,
+                            reasoning_buffer,
                         ):
                             continue
 
@@ -1248,6 +1314,7 @@ def agent(
 
                         turn_done.clear()
                         turn_response.clear()
+                        reasoning_buffer.clear()
                         renderer = StreamRenderer(
                             render_markdown=markdown,
                             bot_name=config.agents.defaults.bot_name,

nanobot/cli/onboard.py

@@ -22,7 +22,7 @@ from nanobot.cli.models import (
     get_model_suggestions,
 )
 from nanobot.config.loader import get_config_path, load_config
-from nanobot.config.schema import Config
+from nanobot.config.schema import Config, ModelPresetConfig
 
 console = Console()
 
@@ -49,6 +49,10 @@ _SELECT_FIELD_HINTS: dict[str, tuple[list[str], str]] = {
 
 _BACK_PRESSED = object()  # Sentinel value for back navigation
 
+# Cache of model-preset names populated at runtime so that field handlers can
+# offer existing presets as choices (e.g. AgentDefaults.model_preset).
+_MODEL_PRESET_CACHE: set[str] = set()
+
 
 def _get_questionary():
     """Return questionary or raise a clear error when wizard deps are unavailable."""
@@ -588,9 +592,102 @@ def _handle_context_window_field(
         setattr(working_model, field_name, new_value)
 
 
+def _handle_model_preset_field(
+    working_model: BaseModel, field_name: str, field_display: str, current_value: Any
+) -> None:
+    """Handle the 'model_preset' field with a list of existing presets."""
+    preset_names = sorted(_MODEL_PRESET_CACHE)
+    choices = ["(clear/unset)"] + preset_names
+    default_choice = str(current_value) if current_value else "(clear/unset)"
+    new_value = _select_with_back(field_display, choices, default=default_choice)
+    if new_value is _BACK_PRESSED:
+        return
+    if new_value == "(clear/unset)":
+        setattr(working_model, field_name, None)
+    elif new_value is not None:
+        setattr(working_model, field_name, new_value)
+
+
+def _handle_provider_field(
+    working_model: BaseModel, field_name: str, field_display: str, current_value: Any
+) -> None:
+    """Handle the 'provider' field with a list of registered providers."""
+    provider_names = sorted(_get_provider_names().keys())
+    choices = ["auto"] + provider_names
+    default_choice = str(current_value) if current_value else "auto"
+    new_value = _select_with_back(field_display, choices, default=default_choice)
+    if new_value is _BACK_PRESSED:
+        return
+    if new_value is not None:
+        setattr(working_model, field_name, new_value)
+
+
+def _handle_fallback_models_field(
+    working_model: BaseModel, field_name: str, field_display: str, current_value: Any
+) -> None:
+    """Handle the 'fallback_models' field with preset-aware list management."""
+    from nanobot.config.schema import InlineFallbackConfig
+
+    items: list[Any] = list(current_value) if isinstance(current_value, list) else []
+    preset_names = sorted(_MODEL_PRESET_CACHE)
+
+    while True:
+        console.clear()
+        console.print(f"[bold]{field_display}[/bold]")
+        if items:
+            for idx, item in enumerate(items, 1):
+                if isinstance(item, InlineFallbackConfig):
+                    console.print(f"  {idx}. {item.model} ({item.provider}) [inline]")
+                else:
+                    console.print(f"  {idx}. {item}")
+        else:
+            console.print("  [dim](empty)[/dim]")
+        console.print()
+
+        choices = ["[+] Add preset"]
+        if items:
+            choices.append("[-] Remove last")
+            choices.append("[X] Clear all")
+        choices.append("[Done]")
+        choices.append("<- Back")
+
+        answer = _get_questionary().select(
+            "Manage fallback models:",
+            choices=choices,
+            qmark=">",
+        ).ask()
+
+        if answer is None or answer == "<- Back":
+            return
+        if answer == "[Done]":
+            setattr(working_model, field_name, items)
+            return
+        if answer == "[+] Add preset":
+            if not preset_names:
+                console.print("[yellow]! No presets defined yet.[/yellow]")
+                _get_questionary().press_any_key_to_continue().ask()
+                continue
+            add_choices = [p for p in preset_names if p not in items]
+            if not add_choices:
+                console.print("[yellow]! All presets already added.[/yellow]")
+                _get_questionary().press_any_key_to_continue().ask()
+                continue
+            picked = _select_with_back("Select preset:", add_choices)
+            if picked is _BACK_PRESSED or picked is None:
+                continue
+            items.append(picked)
+        elif answer == "[-] Remove last" and items:
+            items.pop()
+        elif answer == "[X] Clear all" and items:
+            items.clear()
+
+
 _FIELD_HANDLERS: dict[str, Any] = {
     "model": _handle_model_field,
     "context_window_tokens": _handle_context_window_field,
+    "model_preset": _handle_model_preset_field,
+    "provider": _handle_provider_field,
+    "fallback_models": _handle_fallback_models_field,
 }
 
 
@@ -757,6 +854,116 @@ def _try_auto_fill_context_window(model: BaseModel, new_model_name: str) -> None
         console.print("[dim](i) Could not auto-fill context window (model not in database)[/dim]")
 
 
+# --- Model Preset Configuration ---
+
+
+def _sync_preset_cache(config: Config) -> None:
+    """Synchronise the module-level preset name cache from config."""
+    _MODEL_PRESET_CACHE.clear()
+    _MODEL_PRESET_CACHE.update(config.model_presets.keys())
+
+
+def _configure_model_presets(config: Config) -> None:
+    """Configure model presets (CRUD)."""
+    _sync_preset_cache(config)
+
+    def get_preset_choices() -> list[str]:
+        choices: list[str] = []
+        for name, preset in config.model_presets.items():
+            choices.append(f"{name} ({preset.model})")
+        choices.append("[+] Add new preset")
+        choices.append("<- Back")
+        return choices
+
+    last_preset_name: str | None = None
+    while True:
+        try:
+            console.clear()
+            _show_section_header(
+                "Model Presets",
+                "Create, edit or delete named model presets for quick switching",
+            )
+            choices = get_preset_choices()
+            default_choice = None
+            if last_preset_name:
+                for c in choices:
+                    if c.startswith(last_preset_name + " ("):
+                        default_choice = c
+                        break
+            answer = _select_with_back(
+                "Select preset:", choices, default=default_choice
+            )
+
+            if answer is _BACK_PRESSED or answer is None or answer == "<- Back":
+                break
+
+            assert isinstance(answer, str)
+
+            if answer == "[+] Add new preset":
+                name_input = _get_questionary().text(
+                    "Preset name:",
+                    validate=lambda t: True if t and t.strip() else "Name cannot be empty",
+                ).ask()
+                if not name_input:
+                    continue
+                name = name_input.strip()
+                if name in config.model_presets:
+                    console.print(f"[yellow]! Preset '{name}' already exists[/yellow]")
+                    _pause()
+                    continue
+                if name == "default":
+                    console.print("[yellow]! 'default' is reserved (auto-generated from Agent Settings)[/yellow]")
+                    _pause()
+                    continue
+                new_preset = ModelPresetConfig(model="")
+                updated = _configure_pydantic_model(new_preset, f"New Preset: {name}")
+                if updated is not None:
+                    config.model_presets[name] = updated
+                    _sync_preset_cache(config)
+                    last_preset_name = name
+                continue
+
+            # Editing / deleting an existing preset
+            preset_name = answer.split(" (", 1)[0]
+            preset = config.model_presets.get(preset_name)
+            if preset is None:
+                continue
+
+            last_preset_name = preset_name
+
+            choices = ["Edit", "Cancel"]
+            if preset_name != "default":
+                choices.insert(1, "Delete")
+            action = _select_with_back(
+                f"Preset: {preset_name}",
+                choices,
+                default="Edit",
+            )
+            if action is _BACK_PRESSED or action == "Cancel" or action is None:
+                continue
+
+            if action == "Delete":
+                confirm = _get_questionary().confirm(
+                    f"Delete preset '{preset_name}'?",
+                    default=False,
+                ).ask()
+                if confirm:
+                    del config.model_presets[preset_name]
+                    _sync_preset_cache(config)
+                    last_preset_name = None
+                continue
+
+            if action == "Edit":
+                updated = _configure_pydantic_model(preset, f"Edit Preset: {preset_name}")
+                if updated is not None:
+                    config.model_presets[preset_name] = updated
+                    _sync_preset_cache(config)
+
+        except KeyboardInterrupt:
+            console.print("\n[dim]Returning to main menu...[/dim]")
+            break
+
+
 # --- Provider Configuration ---
 
 
@@ -1043,6 +1250,12 @@ def _show_summary(config: Config) -> None:
         channel_rows.append((display, status))
     _print_summary_panel(channel_rows, "Chat Channels")
 
+    # Model Presets
+    preset_rows = []
+    for name, preset in config.model_presets.items():
+        preset_rows.append((name, f"{preset.model} (ctx={preset.context_window_tokens})"))
+    _print_summary_panel(preset_rows, "Model Presets")
+
     # Settings sections
     for title, model in [
         ("Agent Settings", config.agents.defaults),
@@ -1112,6 +1325,7 @@ def run_onboard(initial_config: Config | None = None) -> OnboardResult:
 
     original_config = base_config.model_copy(deep=True)
     config = base_config.model_copy(deep=True)
+    _sync_preset_cache(config)
 
     last_main_choice: str | None = None
     while True:
@@ -1123,6 +1337,7 @@ def run_onboard(initial_config: Config | None = None) -> OnboardResult:
                 "What would you like to configure?",
                 choices=[
                     "[P] LLM Provider",
+                    "[M] Model Presets",
                     "[C] Chat Channel",
                     "[H] Channel Common",
                     "[A] Agent Settings",
@@ -1149,6 +1364,7 @@ def run_onboard(initial_config: Config | None = None) -> OnboardResult:
 
         _menu_dispatch = {
             "[P] LLM Provider": lambda: _configure_providers(config),
+            "[M] Model Presets": lambda: _configure_model_presets(config),
             "[C] Chat Channel": lambda: _configure_channels(config),
             "[H] Channel Common": lambda: _configure_general_settings(config, "Channel Common"),
             "[A] Agent Settings": lambda: _configure_general_settings(config, "Agent Settings"),

nanobot/config/schema.py

@@ -155,8 +155,35 @@ class AgentDefaults(Base):
         validation_alias=AliasChoices("consolidationRatio"),
         serialization_alias="consolidationRatio",
     )  # Consolidation target ratio (0.5 = 50% of budget retained after compression)
+    vision_models: list[str] = Field(default_factory=list)  # Models that support image input
+    audio_models: list[str] = Field(default_factory=list)  # Models that support native audio input
+    video_models: list[str] = Field(default_factory=list)  # Models that support native video input
     dream: DreamConfig = Field(default_factory=DreamConfig)
 
+    @staticmethod
+    def _bare_model(model: str) -> str:
+        """Strip provider prefix, e.g. 'openai/gpt-4o' -> 'gpt-4o'."""
+        return model.split("/", 1)[-1].lower() if "/" in model else model.lower()
+
+    def _supports_capability(self, model: str, patterns: list[str]) -> bool | None:
+        """Check if model matches any pattern. Returns None if patterns is empty."""
+        if not patterns:
+            return None
+        bare = self._bare_model(model)
+        return any(p.lower() in bare for p in patterns)
+
+    def supports_vision(self, model: str) -> bool | None:
+        """Check if model supports vision. None if unconfigured."""
+        return self._supports_capability(model, self.vision_models)
+
+    def supports_audio(self, model: str) -> bool | None:
+        """Check if model supports native audio. None if unconfigured."""
+        return self._supports_capability(model, self.audio_models)
+
+    def supports_video(self, model: str) -> bool | None:
+        """Check if model supports native video. None if unconfigured."""
+        return self._supports_capability(model, self.video_models)
+
 
 class AgentsConfig(Base):
     """Agent configuration."""
@@ -190,6 +217,7 @@ class ProvidersConfig(Base):
     openai: ProviderConfig = Field(default_factory=ProviderConfig)
     openrouter: ProviderConfig = Field(default_factory=ProviderConfig)
     huggingface: ProviderConfig = Field(default_factory=ProviderConfig)
+    skywork: ProviderConfig = Field(default_factory=ProviderConfig)  # Skywork / APIFree API gateway
     deepseek: ProviderConfig = Field(default_factory=ProviderConfig)
     groq: ProviderConfig = Field(default_factory=ProviderConfig)
     zhipu: ProviderConfig = Field(default_factory=ProviderConfig)
@@ -207,6 +235,7 @@ class ProvidersConfig(Base):
     stepfun: ProviderConfig = Field(default_factory=ProviderConfig)  # Step Fun (阶跃星辰)
     xiaomi_mimo: ProviderConfig = Field(default_factory=ProviderConfig)  # Xiaomi MIMO (小米)
     longcat: ProviderConfig = Field(default_factory=ProviderConfig)  # LongCat
+    ant_ling: ProviderConfig = Field(default_factory=ProviderConfig)  # Ant Ling
     aihubmix: ProviderConfig = Field(default_factory=ProviderConfig)  # AiHubMix API gateway
     siliconflow: ProviderConfig = Field(default_factory=ProviderConfig)  # SiliconFlow (硅基流动)
     volcengine: ProviderConfig = Field(default_factory=ProviderConfig)  # VolcEngine (火山引擎)
@@ -256,6 +285,17 @@ class MCPServerConfig(Base):
     enabled_tools: list[str] = Field(default_factory=lambda: ["*"])  # Only register these tools; accepts raw MCP names or wrapped mcp_<server>_<tool> names; ["*"] = all tools; [] = no tools
 
 
+class InputLimitsConfig(Base):
+    """Limits for user-provided multimodal inputs."""
+
+    max_input_images: int = 3
+    max_input_image_bytes: int = 10 * 1024 * 1024  # 10 MB
+    max_input_audios: int = 1
+    max_input_audio_bytes: int = 10 * 1024 * 1024  # 10 MB
+    max_input_videos: int = 1
+    max_input_video_bytes: int = 20 * 1024 * 1024  # 20 MB
+
+
 def _lazy_default(module_path: str, class_name: str) -> Any:
     """Deferred import helper for ToolsConfig default factories."""
     import importlib
@@ -277,6 +317,7 @@ class ToolsConfig(Base):
     image_generation: ImageGenerationToolConfig = Field(
         default_factory=lambda: _lazy_default("nanobot.agent.tools.image_generation", "ImageGenerationToolConfig"),
     )
+    input_limits: InputLimitsConfig = Field(default_factory=InputLimitsConfig)
     restrict_to_workspace: bool = False  # restrict all tool access to workspace directory
     mcp_servers: dict[str, MCPServerConfig] = Field(default_factory=dict)
     ssrf_whitelist: list[str] = Field(default_factory=list)  # CIDR ranges to exempt from SSRF blocking (e.g. ["100.64.0.0/10"] for Tailscale)

nanobot/heartbeat/service.py

@@ -4,12 +4,12 @@ from __future__ import annotations
 
 import asyncio
 from pathlib import Path
-from typing import TYPE_CHECKING, Any, Callable, Coroutine
+from typing import Any, Callable, Coroutine
 
 from loguru import logger
 
-if TYPE_CHECKING:
-    from nanobot.providers.base import LLMProvider
+from nanobot.providers.base import LLMProvider
+from nanobot.utils.llm_runtime import LLMRuntimeResolver, static_llm_runtime
 
 _HEARTBEAT_TOOL = [
     {
@@ -53,17 +53,21 @@ class HeartbeatService:
     def __init__(
         self,
         workspace: Path,
-        provider: LLMProvider,
-        model: str,
+        provider: LLMProvider | None = None,
+        model: str | None = None,
         on_execute: Callable[[str], Coroutine[Any, Any, str]] | None = None,
         on_notify: Callable[[str], Coroutine[Any, Any, None]] | None = None,
         interval_s: int = 30 * 60,
         enabled: bool = True,
         timezone: str | None = None,
+        llm_runtime: LLMRuntimeResolver | None = None,
     ):
         self.workspace = workspace
-        self.provider = provider
-        self.model = model
+        if llm_runtime is None:
+            if provider is None or model is None:
+                raise ValueError("HeartbeatService requires either llm_runtime or provider/model")
+            llm_runtime = static_llm_runtime(provider, model)
+        self._llm_runtime = llm_runtime
         self.on_execute = on_execute
         self.on_notify = on_notify
         self.interval_s = interval_s
@@ -91,7 +95,9 @@ class HeartbeatService:
         """
         from nanobot.utils.helpers import current_time_str
 
-        response = await self.provider.chat_with_retry(
+        llm = self._llm_runtime()
+
+        response = await llm.provider.chat_with_retry(
             messages=[
                 {"role": "system", "content": "You are a heartbeat agent. Call the heartbeat tool to report your decision."},
                 {"role": "user", "content": (
@@ -101,7 +107,7 @@ class HeartbeatService:
                 )},
             ],
             tools=_HEARTBEAT_TOOL,
-            model=self.model,
+            model=llm.model,
         )
 
         if not response.should_execute_tools:
@@ -214,8 +220,9 @@ class HeartbeatService:
                     )
                     return
 
+                llm = self._llm_runtime()
                 should_notify = await evaluate_response(
-                    response, tasks, self.provider, self.model,
+                    response, tasks, llm.provider, llm.model,
                 )
                 if should_notify and self.on_notify:
                     logger.info("Heartbeat: completed, delivering response")

nanobot/nanobot.py

@@ -8,6 +8,7 @@ from typing import Any
 
 from nanobot.agent.hook import AgentHook, SDKCaptureHook
 from nanobot.agent.loop import AgentLoop
+from nanobot.providers.image_generation import image_gen_provider_configs
 
 
 @dataclass(slots=True)
@@ -63,10 +64,7 @@ class Nanobot:
 
         loop = AgentLoop.from_config(
             config,
-            image_generation_provider_configs={
-                "openrouter": config.providers.openrouter,
-                "aihubmix": config.providers.aihubmix,
-            },
+            image_generation_provider_configs=image_gen_provider_configs(config),
         )
         return cls(loop)
 

nanobot/providers/anthropic_provider.py

@@ -212,7 +212,7 @@ class AnthropicProvider(LLMProvider):
 
     @staticmethod
     def _convert_user_content(content: Any) -> Any:
-        """Convert user message content, translating image_url blocks."""
+        """Convert user message content, translating image_url and input_audio blocks."""
         if isinstance(content, str) or content is None:
             return content or "(empty)"
         if not isinstance(content, list):
@@ -228,6 +228,14 @@ class AnthropicProvider(LLMProvider):
                 if converted:
                     result.append(converted)
                 continue
+            if item.get("type") == "input_audio":
+                # Anthropic doesn't support native audio → text placeholder
+                result.append(LLMProvider._media_placeholder("input_audio", item))
+                continue
+            if item.get("type") == "video_url":
+                # Anthropic doesn't support native video → text placeholder
+                result.append(LLMProvider._media_placeholder("video_url", item))
+                continue
             result.append(item)
         return result or "(empty)"
 
@@ -590,6 +598,7 @@ class AnthropicProvider(LLMProvider):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         kwargs = self._build_kwargs(
             messages, tools, model, max_tokens, temperature,
@@ -598,11 +607,12 @@ class AnthropicProvider(LLMProvider):
         idle_timeout_s = int(os.environ.get("NANOBOT_STREAM_IDLE_TIMEOUT_S", "90"))
         try:
             async with self._client.messages.stream(**kwargs) as stream:
-                if on_content_delta or on_thinking_delta:
+                if on_content_delta or on_thinking_delta or on_tool_call_delta:
                     # Idle timeout must track *any* SSE chunk (thinking_delta,
                     # tool JSON deltas, etc.), not only text_stream tokens.
                     # Otherwise extended thinking can stall text_stream for minutes
                     # while the connection is healthy (e.g. MiniMax Anthropic).
+                    tool_blocks: dict[int, dict[str, str]] = {}
                     while True:
                         try:
                             chunk = await asyncio.wait_for(
@@ -611,7 +621,22 @@ class AnthropicProvider(LLMProvider):
                             )
                         except StopAsyncIteration:
                             break
-                        if (
+                        if chunk.type == "content_block_start":
+                            block = getattr(chunk, "content_block", None)
+                            if getattr(block, "type", None) == "tool_use":
+                                index = int(getattr(chunk, "index", 0) or 0)
+                                state = {
+                                    "call_id": str(getattr(block, "id", "") or ""),
+                                    "name": str(getattr(block, "name", "") or ""),
+                                }
+                                tool_blocks[index] = state
+                                if on_tool_call_delta:
+                                    await on_tool_call_delta({
+                                        "index": index,
+                                        **state,
+                                        "arguments_delta": "",
+                                    })
+                        elif (
                             chunk.type == "content_block_delta"
                             and getattr(chunk.delta, "type", None) == "thinking_delta"
                         ):
@@ -625,6 +650,20 @@ class AnthropicProvider(LLMProvider):
                             text = getattr(chunk.delta, "text", None) or ""
                             if text and on_content_delta:
                                 await on_content_delta(text)
+                        elif (
+                            chunk.type == "content_block_delta"
+                            and getattr(chunk.delta, "type", None) == "input_json_delta"
+                        ):
+                            partial = getattr(chunk.delta, "partial_json", None) or ""
+                            if partial and on_tool_call_delta:
+                                index = int(getattr(chunk, "index", 0) or 0)
+                                state = tool_blocks.get(index, {})
+                                await on_tool_call_delta({
+                                    "index": index,
+                                    "call_id": state.get("call_id", ""),
+                                    "name": state.get("name", ""),
+                                    "arguments_delta": partial,
+                                })
                 response = await asyncio.wait_for(
                     stream.get_final_message(),
                     timeout=idle_timeout_s,

nanobot/providers/azure_openai_provider.py

@@ -158,6 +158,7 @@ class AzureOpenAIProvider(LLMProvider):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         _ = on_thinking_delta
         body = self._build_body(
@@ -169,7 +170,7 @@ class AzureOpenAIProvider(LLMProvider):
         try:
             stream = await self._client.responses.create(**body)
             content, tool_calls, finish_reason, usage, reasoning_content = (
-                await consume_sdk_stream(stream, on_content_delta)
+                await consume_sdk_stream(stream, on_content_delta, on_tool_call_delta)
             )
             return LLMResponse(
                 content=content or None,

nanobot/providers/base.py

@@ -13,8 +13,6 @@ from typing import Any
 
 from loguru import logger
 
-from nanobot.utils.helpers import image_placeholder_text
-
 
 @dataclass
 class ToolCallRequest:
@@ -70,11 +68,11 @@ class LLMResponse:
 
     @property
     def should_execute_tools(self) -> bool:
-        """Tools execute only when has_tool_calls AND finish_reason is ``tool_calls`` / ``stop``.
+        """Tools execute only when has_tool_calls AND finish_reason is a tool-capable stop.
         Blocks gateway-injected calls under ``refusal`` / ``content_filter`` / ``error`` (#3220)."""
         if not self.has_tool_calls:
             return False
-        return self.finish_reason in ("tool_calls", "stop")
+        return self.finish_reason in ("tool_calls", "function_call", "stop")
 
 
 @dataclass(frozen=True)
@@ -112,6 +110,7 @@ class LLMProvider(ABC):
         "server error",
         "temporarily unavailable",
         "速率限制",
+        "访问量过大",
     )
     _RETRYABLE_STATUS_CODES = frozenset({408, 409, 429})
     _TRANSIENT_ERROR_KINDS = frozenset({"timeout", "connection"})
@@ -438,9 +437,23 @@ class LLMProvider(ABC):
 
         return merged
 
+    _MEDIA_LABEL_MAP = {"image_url": "image", "input_audio": "audio", "video_url": "video"}
+    _STRIP_MEDIA_TYPES = frozenset({"image_url", "input_audio", "video_url"})
+
     @staticmethod
-    def _strip_image_content(messages: list[dict[str, Any]]) -> list[dict[str, Any]] | None:
-        """Replace image_url blocks with text placeholder. Returns None if no images found."""
+    def _media_placeholder(btype: str, block: dict[str, Any]) -> dict[str, str]:
+        """Build a text placeholder for a media block."""
+        path = (block.get("_meta") or {}).get("path", "")
+        label = LLMProvider._MEDIA_LABEL_MAP.get(btype, "media")
+        text = f"[{label}: {path}]" if path else f"[{label}]"
+        return {"type": "text", "text": text}
+
+    @staticmethod
+    def _strip_media_content(messages: list[dict[str, Any]]) -> list[dict[str, Any]] | None:
+        """Replace image_url, input_audio, and video_url blocks with text placeholders.
+
+        Returns None if no media blocks were found (no changes needed).
+        """
         found = False
         result = []
         for msg in messages:
@@ -448,10 +461,8 @@ class LLMProvider(ABC):
             if isinstance(content, list):
                 new_content = []
                 for b in content:
-                    if isinstance(b, dict) and b.get("type") == "image_url":
-                        path = (b.get("_meta") or {}).get("path", "")
-                        placeholder = image_placeholder_text(path, empty="[image omitted]")
-                        new_content.append({"type": "text", "text": placeholder})
+                    if isinstance(b, dict) and b.get("type") in LLMProvider._STRIP_MEDIA_TYPES:
+                        new_content.append(LLMProvider._media_placeholder(b["type"], b))
                         found = True
                     else:
                         new_content.append(b)
@@ -461,8 +472,13 @@ class LLMProvider(ABC):
         return result if found else None
 
     @staticmethod
-    def _strip_image_content_inplace(messages: list[dict[str, Any]]) -> bool:
-        """Replace image_url blocks with text placeholder *in-place*.
+    def _strip_image_content(messages: list[dict[str, Any]]) -> list[dict[str, Any]] | None:
+        """Replace image_url blocks with text placeholder. Returns None if no images found."""
+        return LLMProvider._strip_media_content(messages)
+
+    @staticmethod
+    def _strip_media_content_inplace(messages: list[dict[str, Any]]) -> bool:
+        """Replace media blocks with text placeholder *in-place*.
 
         Mutates the content lists of the original message dicts so that
         callers holding references to those dicts also see the stripped
@@ -473,13 +489,16 @@ class LLMProvider(ABC):
             content = msg.get("content")
             if isinstance(content, list):
                 for i, b in enumerate(content):
-                    if isinstance(b, dict) and b.get("type") == "image_url":
-                        path = (b.get("_meta") or {}).get("path", "")
-                        placeholder = image_placeholder_text(path, empty="[image omitted]")
-                        content[i] = {"type": "text", "text": placeholder}
+                    if isinstance(b, dict) and b.get("type") in LLMProvider._STRIP_MEDIA_TYPES:
+                        content[i] = LLMProvider._media_placeholder(b["type"], b)
                         found = True
         return found
 
+    @staticmethod
+    def _strip_image_content_inplace(messages: list[dict[str, Any]]) -> bool:
+        """Replace image_url blocks with text placeholder *in-place*."""
+        return LLMProvider._strip_media_content_inplace(messages)
+
     async def _safe_chat(self, **kwargs: Any) -> LLMResponse:
         """Call chat() and convert unexpected exceptions to error responses."""
         try:
@@ -500,6 +519,7 @@ class LLMProvider(ABC):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         """Stream a chat completion, calling *on_content_delta* for each text chunk.
 
@@ -513,7 +533,7 @@ class LLMProvider(ABC):
         full content as a single delta.  Providers that support native
         streaming should override this method.
         """
-        _ = on_thinking_delta
+        _ = on_thinking_delta, on_tool_call_delta
         response = await self.chat(
             messages=messages, tools=tools, model=model,
             max_tokens=max_tokens, temperature=temperature,
@@ -543,6 +563,7 @@ class LLMProvider(ABC):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
         retry_mode: str = "standard",
         on_retry_wait: Callable[[str], Awaitable[None]] | None = None,
     ) -> LLMResponse:
@@ -560,6 +581,7 @@ class LLMProvider(ABC):
             reasoning_effort=reasoning_effort, tool_choice=tool_choice,
             on_content_delta=on_content_delta,
             on_thinking_delta=on_thinking_delta,
+            on_tool_call_delta=on_tool_call_delta,
         )
         return await self._run_with_retry(
             self._safe_chat_stream,
@@ -734,18 +756,18 @@ class LLMProvider(ABC):
                 identical_error_count = 1 if error_key else 0
 
             if not self._is_transient_response(response):
-                stripped = self._strip_image_content(original_messages)
+                stripped = self._strip_media_content(original_messages)
                 if stripped is not None and stripped != kw["messages"]:
                     logger.warning(
-                        "Non-transient LLM error with image content, retrying without images"
+                        "Non-transient LLM error with media content, retrying without media"
                     )
                     retry_kw = dict(kw)
                     retry_kw["messages"] = stripped
                     result = await call(**retry_kw)
-                    # Permanently strip images from the original messages so
+                    # Permanently strip media from the original messages so
                     # subsequent iterations do not repeat the error-retry cycle.
                     if result.finish_reason != "error":
-                        self._strip_image_content_inplace(original_messages)
+                        self._strip_media_content_inplace(original_messages)
                     return result
                 return response
 

nanobot/providers/bedrock_provider.py

@@ -704,8 +704,9 @@ class BedrockProvider(LLMProvider):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
-        _ = on_thinking_delta
+        _ = on_thinking_delta, on_tool_call_delta
         idle_timeout_s = int(os.environ.get("NANOBOT_STREAM_IDLE_TIMEOUT_S", "90"))
         content_parts: list[str] = []
         reasoning_parts: list[str] = []

nanobot/providers/github_copilot_provider.py

@@ -243,6 +243,7 @@ class GitHubCopilotProvider(OpenAICompatProvider):
         tool_choice: str | dict[str, object] | None = None,
         on_content_delta: Callable[[str], None] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, object]], Awaitable[None]] | None = None,
     ):
         await self._refresh_client_api_key()
         return await super().chat_stream(
@@ -255,4 +256,5 @@ class GitHubCopilotProvider(OpenAICompatProvider):
             tool_choice=tool_choice,
             on_content_delta=on_content_delta,
             on_thinking_delta=on_thinking_delta,
+            on_tool_call_delta=on_tool_call_delta,
         )

nanobot/providers/image_generation.py

@@ -3,11 +3,14 @@
 from __future__ import annotations
 
 import base64
+import binascii
+from abc import ABC, abstractmethod
 from dataclasses import dataclass
 from pathlib import Path
 from typing import Any
 
 import httpx
+from loguru import logger
 
 from nanobot.providers.registry import find_by_name
 from nanobot.utils.helpers import detect_image_mime
@@ -26,6 +29,8 @@ _AIHUBMIX_ASPECT_RATIO_SIZES = {
     "4:3": "1536x1024",
     "16:9": "1536x1024",
 }
+_GEMINI_DEFAULT_TIMEOUT_S = 120.0
+_GEMINI_IMAGEN_ASPECT_RATIOS = {"1:1", "9:16", "16:9", "3:4", "4:3"}
 
 
 class ImageGenerationError(RuntimeError):
@@ -41,28 +46,38 @@ class GeneratedImageResponse:
     raw: dict[str, Any]
 
 
-def _provider_base_url(provider: str, api_base: str | None, fallback: str) -> str:
-    if api_base:
-        return api_base.rstrip("/")
-    spec = find_by_name(provider)
-    if spec and spec.default_api_base:
-        return spec.default_api_base.rstrip("/")
-    return fallback
-
-
-def image_path_to_data_url(path: str | Path) -> str:
-    """Convert a local image path to an image data URL."""
+def _read_image_b64(path: str | Path) -> tuple[str, str]:
+    """Return ``(mime, base64)`` for the image at ``path``."""
     p = Path(path).expanduser()
     raw = p.read_bytes()
     mime = detect_image_mime(raw)
     if mime is None:
         raise ImageGenerationError(f"unsupported reference image: {p}")
-    encoded = base64.b64encode(raw).decode("ascii")
+    return mime, base64.b64encode(raw).decode("ascii")
+
+
+def image_path_to_data_url(path: str | Path) -> str:
+    """Convert a local image path to an image data URL."""
+    mime, encoded = _read_image_b64(path)
     return f"data:{mime};base64,{encoded}"
 
 
-def _b64_png_data_url(value: str) -> str:
-    return f"data:image/png;base64,{value}"
+def image_path_to_inline_data(path: str | Path) -> dict[str, str]:
+    """Convert a local image path to a Gemini ``inlineData`` payload dict."""
+    mime, encoded = _read_image_b64(path)
+    return {"mimeType": mime, "data": encoded}
+
+
+def _b64_image_data_url(value: str) -> str:
+    encoded = "".join(value.split())
+    try:
+        raw = base64.b64decode(encoded, validate=True)
+    except binascii.Error as exc:
+        raise ImageGenerationError("generated image payload was not valid base64") from exc
+    mime = detect_image_mime(raw)
+    if mime is None:
+        raise ImageGenerationError("generated image payload was not a supported image")
+    return f"data:{mime};base64,{encoded}"
 
 
 def _aihubmix_size(aspect_ratio: str | None, image_size: str | None) -> str:
@@ -106,8 +121,49 @@ async def _download_image_data_url(
     return f"data:{mime};base64,{encoded}"
 
 
-class OpenRouterImageGenerationClient:
-    """Small async client for OpenRouter Chat Completions image generation."""
+# ---------------------------------------------------------------------------
+# Registry
+# ---------------------------------------------------------------------------
+
+_IMAGE_GEN_PROVIDERS: dict[str, type[ImageGenerationProvider]] = {}
+
+
+def register_image_gen_provider(cls: type[ImageGenerationProvider]) -> None:
+    name = cls.provider_name
+    if not name:
+        raise ValueError(f"{cls.__name__} must set provider_name")
+    _IMAGE_GEN_PROVIDERS[name] = cls
+
+
+def get_image_gen_provider(name: str) -> type[ImageGenerationProvider] | None:
+    return _IMAGE_GEN_PROVIDERS.get(name)
+
+
+def image_gen_provider_names() -> tuple[str, ...]:
+    """Return registered image generation provider names in registry order."""
+    return tuple(_IMAGE_GEN_PROVIDERS)
+
+
+def image_gen_provider_configs(config: Any) -> dict[str, Any]:
+    providers_cfg = config.providers
+    return {
+        name: pc
+        for name in _IMAGE_GEN_PROVIDERS
+        if (pc := getattr(providers_cfg, name, None)) is not None
+    }
+
+
+# ---------------------------------------------------------------------------
+# Base class
+# ---------------------------------------------------------------------------
+
+
+class ImageGenerationProvider(ABC):
+    """Base class for image generation provider clients."""
+
+    provider_name: str = ""
+    missing_key_message: str = ""
+    default_timeout: float = _DEFAULT_TIMEOUT_S
 
     def __init__(
         self,
@@ -116,20 +172,71 @@ class OpenRouterImageGenerationClient:
         api_base: str | None = None,
         extra_headers: dict[str, str] | None = None,
         extra_body: dict[str, Any] | None = None,
-        timeout: float = _DEFAULT_TIMEOUT_S,
+        timeout: float | None = None,
         client: httpx.AsyncClient | None = None,
     ) -> None:
         self.api_key = api_key
-        self.api_base = _provider_base_url(
-            "openrouter",
-            api_base,
-            "https://openrouter.ai/api/v1",
-        )
+        self.api_base = self._resolve_base_url(api_base)
         self.extra_headers = extra_headers or {}
         self.extra_body = extra_body or {}
-        self.timeout = timeout
+        self.timeout = timeout if timeout is not None else self.default_timeout
         self._client = client
 
+    def _resolve_base_url(self, api_base: str | None) -> str:
+        if api_base:
+            return api_base.rstrip("/")
+        spec = find_by_name(self.provider_name)
+        if spec and spec.default_api_base:
+            return spec.default_api_base.rstrip("/")
+        return self._default_base_url()
+
+    def _default_base_url(self) -> str:
+        return ""
+
+    @abstractmethod
+    async def generate(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        reference_images: list[str] | None = None,
+        aspect_ratio: str | None = None,
+        image_size: str | None = None,
+    ) -> GeneratedImageResponse: ...
+
+    def _require_images(self, images: list[str], data: dict[str, Any]) -> None:
+        if images:
+            return
+        provider_error = data.get("error") if isinstance(data, dict) else None
+        label = self.provider_name
+        if provider_error:
+            raise ImageGenerationError(f"{label} returned no images: {provider_error}")
+        raise ImageGenerationError(f"{label} returned no images for this request")
+
+    async def _http_post(
+        self,
+        url: str,
+        *,
+        headers: dict[str, str],
+        body: dict[str, Any],
+    ) -> httpx.Response:
+        if self._client is not None:
+            return await self._client.post(url, headers=headers, json=body)
+        async with httpx.AsyncClient(timeout=self.timeout) as c:
+            return await c.post(url, headers=headers, json=body)
+
+
+class OpenRouterImageGenerationClient(ImageGenerationProvider):
+    """Small async client for OpenRouter Chat Completions image generation."""
+
+    provider_name = "openrouter"
+    missing_key_message = (
+        "OpenRouter API key is not configured. Set providers.openrouter.apiKey."
+    )
+
+    def _default_base_url(self) -> str:
+        return "https://openrouter.ai/api/v1"
+
     async def generate(
         self,
         *,
@@ -140,9 +247,7 @@ class OpenRouterImageGenerationClient:
         image_size: str | None = None,
     ) -> GeneratedImageResponse:
         if not self.api_key:
-            raise ImageGenerationError(
-                "OpenRouter API key is not configured. Set providers.openrouter.apiKey."
-            )
+            raise ImageGenerationError(self.missing_key_message)
 
         content: str | list[dict[str, Any]]
         references = list(reference_images or [])
@@ -178,12 +283,7 @@ class OpenRouterImageGenerationClient:
             **self.extra_headers,
         }
         url = f"{self.api_base}/chat/completions"
-
-        if self._client is not None:
-            response = await self._client.post(url, headers=headers, json=body)
-        else:
-            async with httpx.AsyncClient(timeout=self.timeout) as client:
-                response = await client.post(url, headers=headers, json=body)
+        response = await self._http_post(url, headers=headers, body=body)
 
         try:
             response.raise_for_status()
@@ -208,11 +308,7 @@ class OpenRouterImageGenerationClient:
                 if isinstance(url_value, str) and url_value.startswith("data:image/"):
                     images.append(url_value)
 
-        if not images:
-            provider_error = data.get("error") if isinstance(data, dict) else None
-            if provider_error:
-                raise ImageGenerationError(f"OpenRouter returned no images: {provider_error}")
-            raise ImageGenerationError("OpenRouter returned no images for this request")
+        self._require_images(images, data)
 
         return GeneratedImageResponse(
             images=images,
@@ -221,29 +317,17 @@ class OpenRouterImageGenerationClient:
         )
 
 
-class AIHubMixImageGenerationClient:
+class AIHubMixImageGenerationClient(ImageGenerationProvider):
     """Small async client for AIHubMix unified image generation."""
 
-    def __init__(
-        self,
-        *,
-        api_key: str | None,
-        api_base: str | None = None,
-        extra_headers: dict[str, str] | None = None,
-        extra_body: dict[str, Any] | None = None,
-        timeout: float = _AIHUBMIX_TIMEOUT_S,
-        client: httpx.AsyncClient | None = None,
-    ) -> None:
-        self.api_key = api_key
-        self.api_base = _provider_base_url(
-            "aihubmix",
-            api_base,
-            "https://aihubmix.com/v1",
-        )
-        self.extra_headers = extra_headers or {}
-        self.extra_body = extra_body or {}
-        self.timeout = timeout
-        self._client = client
+    provider_name = "aihubmix"
+    missing_key_message = (
+        "AIHubMix API key is not configured. Set providers.aihubmix.apiKey."
+    )
+    default_timeout = _AIHUBMIX_TIMEOUT_S
+
+    def _default_base_url(self) -> str:
+        return "https://aihubmix.com/v1"
 
     async def generate(
         self,
@@ -255,9 +339,7 @@ class AIHubMixImageGenerationClient:
         image_size: str | None = None,
     ) -> GeneratedImageResponse:
         if not self.api_key:
-            raise ImageGenerationError(
-                "AIHubMix API key is not configured. Set providers.aihubmix.apiKey."
-            )
+            raise ImageGenerationError(self.missing_key_message)
 
         refs = list(reference_images or [])
         headers = {
@@ -266,16 +348,8 @@ class AIHubMixImageGenerationClient:
         }
         size = _aihubmix_size(aspect_ratio, image_size)
 
-        if self._client is not None:
-            return await self._generate_with_client(
-                self._client,
-                prompt=prompt,
-                model=model,
-                reference_images=refs,
-                size=size,
-                headers=headers,
-            )
-        async with httpx.AsyncClient(timeout=self.timeout) as client:
+        client = self._client or httpx.AsyncClient(timeout=self.timeout)
+        try:
             return await self._generate_with_client(
                 client,
                 prompt=prompt,
@@ -284,6 +358,9 @@ class AIHubMixImageGenerationClient:
                 size=size,
                 headers=headers,
             )
+        finally:
+            if self._client is None:
+                await client.aclose()
 
     async def _generate_with_client(
         self,
@@ -332,15 +409,182 @@ class AIHubMixImageGenerationClient:
         payload = response.json()
         images = await _aihubmix_images_from_payload(client, payload)
 
-        if not images:
-            provider_error = payload.get("error") if isinstance(payload, dict) else None
-            if provider_error:
-                raise ImageGenerationError(f"AIHubMix returned no images: {provider_error}")
-            raise ImageGenerationError("AIHubMix returned no images for this request")
+        self._require_images(images, payload)
 
         return GeneratedImageResponse(images=images, content="", raw=payload)
 
 
+def _http_error_detail(response: httpx.Response) -> str:
+    """Extract a readable error message from an HTTP error response."""
+    try:
+        data = response.json()
+        if isinstance(data, dict):
+            err = data.get("error")
+            if isinstance(err, dict):
+                return err.get("message") or str(err)
+            if err:
+                return str(err)
+    except Exception:
+        pass
+    return response.text[:500] or "<empty response body>"
+
+
+class GeminiImageGenerationClient(ImageGenerationProvider):
+    """Async client for Gemini/Imagen image generation via the Generative Language API."""
+
+    provider_name = "gemini"
+    missing_key_message = (
+        "Gemini API key is not configured. Set providers.gemini.apiKey."
+    )
+    default_timeout = _GEMINI_DEFAULT_TIMEOUT_S
+
+    def _default_base_url(self) -> str:
+        return "https://generativelanguage.googleapis.com/v1beta"
+
+    def _resolve_base_url(self, api_base: str | None) -> str:
+        # The Gemini provider's registry default_api_base is the OpenAI-compat
+        # shim (.../v1beta/openai/), which has no image endpoints.
+        # Skip the registry lookup and use the native API base directly.
+        if api_base:
+            return api_base.rstrip("/")
+        return self._default_base_url()
+
+    async def generate(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        reference_images: list[str] | None = None,
+        aspect_ratio: str | None = None,
+        image_size: str | None = None,
+    ) -> GeneratedImageResponse:
+        if not self.api_key:
+            raise ImageGenerationError(self.missing_key_message)
+        if "imagen" in model.lower():
+            if reference_images:
+                logger.warning(
+                    "Imagen models do not support reference images; "
+                    "ignoring {} reference image(s) for {}",
+                    len(reference_images),
+                    model,
+                )
+            return await self._generate_imagen(
+                prompt=prompt, model=model, aspect_ratio=aspect_ratio
+            )
+        return await self._generate_gemini_flash(
+            prompt=prompt, model=model, reference_images=reference_images or []
+        )
+
+    async def _generate_imagen(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        aspect_ratio: str | None,
+    ) -> GeneratedImageResponse:
+        parameters: dict[str, Any] = {"sampleCount": 1}
+        if aspect_ratio in _GEMINI_IMAGEN_ASPECT_RATIOS:
+            parameters["aspectRatio"] = aspect_ratio
+        body: dict[str, Any] = {
+            "instances": [{"prompt": prompt}],
+            "parameters": parameters,
+        }
+        body.update(self.extra_body)
+
+        url = f"{self.api_base}/models/{model}:predict"
+        headers = {
+            "x-goog-api-key": self.api_key or "",
+            "Content-Type": "application/json",
+            **self.extra_headers,
+        }
+        response = await self._http_post(url, headers=headers, body=body)
+
+        try:
+            response.raise_for_status()
+        except httpx.HTTPStatusError as exc:
+            detail = _http_error_detail(response)
+            logger.error("Gemini Imagen generation failed (HTTP {}): {}", response.status_code, detail)
+            raise ImageGenerationError(
+                f"Gemini Imagen generation failed (HTTP {response.status_code}): {detail}"
+            ) from exc
+
+        data = response.json()
+        images: list[str] = []
+        for prediction in data.get("predictions") or []:
+            if not isinstance(prediction, dict):
+                continue
+            b64 = prediction.get("bytesBase64Encoded")
+            mime = prediction.get("mimeType", "image/png")
+            if isinstance(b64, str) and b64:
+                images.append(f"data:{mime};base64,{b64}")
+
+        self._require_images(images, data)
+
+        return GeneratedImageResponse(images=images, content="", raw=data)
+
+    async def _generate_gemini_flash(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        reference_images: list[str],
+    ) -> GeneratedImageResponse:
+        parts: list[dict[str, Any]] = [
+            {"inlineData": image_path_to_inline_data(path)} for path in reference_images
+        ]
+        parts.append({"text": prompt})
+
+        body: dict[str, Any] = {
+            "contents": [{"role": "user", "parts": parts}],
+            "generationConfig": {"responseModalities": ["TEXT", "IMAGE"]},
+        }
+        body.update(self.extra_body)
+
+        url = f"{self.api_base}/models/{model}:generateContent"
+        headers = {
+            "x-goog-api-key": self.api_key or "",
+            "Content-Type": "application/json",
+            **self.extra_headers,
+        }
+        response = await self._http_post(url, headers=headers, body=body)
+
+        try:
+            response.raise_for_status()
+        except httpx.HTTPStatusError as exc:
+            detail = _http_error_detail(response)
+            logger.error("Gemini image generation failed (HTTP {}): {}", response.status_code, detail)
+            raise ImageGenerationError(
+                f"Gemini image generation failed (HTTP {response.status_code}): {detail}"
+            ) from exc
+
+        data = response.json()
+        images: list[str] = []
+        text_parts: list[str] = []
+        for candidate in data.get("candidates") or []:
+            if not isinstance(candidate, dict):
+                continue
+            content = candidate.get("content") or {}
+            for part in content.get("parts") or []:
+                if not isinstance(part, dict):
+                    continue
+                if "text" in part:
+                    text_parts.append(part["text"])
+                inline = part.get("inlineData")
+                if isinstance(inline, dict):
+                    mime = inline.get("mimeType", "image/png")
+                    b64 = inline.get("data", "")
+                    if b64:
+                        images.append(f"data:{mime};base64,{b64}")
+
+        self._require_images(images, data)
+
+        return GeneratedImageResponse(
+            images=images,
+            content="\n".join(t for t in text_parts if t).strip(),
+            raw=data,
+        )
+
+
 async def _aihubmix_images_from_payload(
     client: httpx.AsyncClient,
     payload: dict[str, Any],
@@ -368,13 +612,13 @@ async def _aihubmix_images_from_payload(
 
         b64_json = value.get("b64_json")
         if isinstance(b64_json, str) and b64_json:
-            images.append(_b64_png_data_url(b64_json))
+            images.append(_b64_image_data_url(b64_json))
         elif b64_json is not None:
             await collect(b64_json)
 
         bytes_base64 = value.get("bytesBase64") or value.get("bytes_base64") or value.get("base64")
         if isinstance(bytes_base64, str) and bytes_base64:
-            images.append(_b64_png_data_url(bytes_base64))
+            images.append(_b64_image_data_url(bytes_base64))
 
         image_url = value.get("image_url") or value.get("imageUrl")
         if isinstance(image_url, dict):
@@ -393,3 +637,254 @@ async def _aihubmix_images_from_payload(
     for candidate in candidates:
         await collect(candidate)
     return images
+
+
+_MINIMAX_TIMEOUT_S = 300.0
+
+_MINIMAX_ASPECT_RATIO_SIZES = {
+    "1:1": "1:1",
+    "16:9": "16:9",
+    "4:3": "4:3",
+    "3:2": "3:2",
+    "2:3": "2:3",
+    "3:4": "3:4",
+    "9:16": "9:16",
+    "21:9": "21:9",
+}
+
+
+class MiniMaxImageGenerationClient(ImageGenerationProvider):
+    """Async client for MiniMax image generation API."""
+
+    provider_name = "minimax"
+    missing_key_message = (
+        "MiniMax API key is not configured. Set providers.minimax.apiKey."
+    )
+    default_timeout = _MINIMAX_TIMEOUT_S
+
+    def _default_base_url(self) -> str:
+        return "https://api.minimaxi.com/v1"
+
+    def _resolve_aspect_ratio(self, aspect_ratio: str | None) -> str:
+        if aspect_ratio and aspect_ratio in _MINIMAX_ASPECT_RATIO_SIZES:
+            return _MINIMAX_ASPECT_RATIO_SIZES[aspect_ratio]
+        return "1:1"
+
+    async def generate(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        reference_images: list[str] | None = None,
+        aspect_ratio: str | None = None,
+        image_size: str | None = None,
+    ) -> GeneratedImageResponse:
+        if not self.api_key:
+            raise ImageGenerationError(self.missing_key_message)
+
+        headers = {
+            "Authorization": f"Bearer {self.api_key}",
+            "Content-Type": "application/json",
+            **self.extra_headers,
+        }
+
+        body: dict[str, Any] = {
+            "model": model,
+            "prompt": prompt,
+            "response_format": "base64",
+        }
+
+        resolved_ratio = self._resolve_aspect_ratio(aspect_ratio)
+        body["aspect_ratio"] = resolved_ratio
+
+        refs = list(reference_images or [])
+        if refs:
+            image_refs = [image_path_to_data_url(path) for path in refs]
+            body["subject_reference"] = [
+                {"type": "character", "image_file": ref} for ref in image_refs
+            ]
+
+        body.update(self.extra_body)
+
+        client = self._client or httpx.AsyncClient(timeout=self.timeout)
+        try:
+            return await self._generate_with_client(client, body, headers)
+        finally:
+            if self._client is None:
+                await client.aclose()
+
+    async def _generate_with_client(
+        self,
+        client: httpx.AsyncClient,
+        body: dict[str, Any],
+        headers: dict[str, str],
+    ) -> GeneratedImageResponse:
+        url = f"{self.api_base}/image_generation"
+        try:
+            response = await client.post(url, headers=headers, json=body)
+        except httpx.TimeoutException as exc:
+            raise ImageGenerationError("MiniMax image generation timed out") from exc
+        except httpx.RequestError as exc:
+            raise ImageGenerationError(f"MiniMax image generation request failed: {exc}") from exc
+
+        try:
+            response.raise_for_status()
+        except httpx.HTTPStatusError as exc:
+            detail = response.text[:500]
+            raise ImageGenerationError(f"MiniMax image generation failed: {detail}") from exc
+
+        payload = response.json()
+        images = _minimax_images_from_payload(payload)
+
+        self._require_images(images, payload)
+
+        return GeneratedImageResponse(images=images, content="", raw=payload)
+
+
+def _minimax_images_from_payload(payload: dict[str, Any]) -> list[str]:
+    """Extract base64 images from MiniMax API response.
+
+    MiniMax returns images in ``data.image_base64`` (list of base64 strings).
+    """
+    images: list[str] = []
+    data = payload.get("data")
+    if not isinstance(data, dict):
+        return images
+    for b64 in data.get("image_base64") or []:
+        if isinstance(b64, str) and b64:
+            images.append(_b64_image_data_url(b64))
+    return images
+
+
+# ---------------------------------------------------------------------------
+# StepFun (阶跃星辰) image generation
+# ---------------------------------------------------------------------------
+
+_STEPFUN_ASPECT_RATIO_SIZES = {
+    "1:1": "1024x1024",
+    "16:9": "1280x800",
+    "9:16": "800x1280",
+    "3:4": "768x1360",
+    "4:3": "1360x768",
+}
+
+
+class StepFunImageGenerationClient(ImageGenerationProvider):
+    """Async client for StepFun (阶跃星辰) image generation.
+
+    Supports:
+    - Text-to-image via step-image-edit-2 (default model)
+    - Reference-image-guided generation via style_reference (step-1x-medium)
+    """
+
+    provider_name = "stepfun"
+    missing_key_message = (
+        "StepFun API key is not configured. Set providers.stepfun.apiKey."
+    )
+    default_timeout = 120.0
+
+    def _default_base_url(self) -> str:
+        return "https://api.stepfun.com/v1"
+
+    async def generate(
+        self,
+        *,
+        prompt: str,
+        model: str,
+        reference_images: list[str] | None = None,
+        aspect_ratio: str | None = None,
+        image_size: str | None = None,
+    ) -> GeneratedImageResponse:
+        if not self.api_key:
+            raise ImageGenerationError(self.missing_key_message)
+
+        headers = {
+            "Authorization": f"Bearer {self.api_key}",
+            "Content-Type": "application/json",
+            **self.extra_headers,
+        }
+
+        body: dict[str, Any] = {
+            "model": model,
+            "prompt": prompt,
+            "response_format": "b64_json",
+            "n": 1,
+        }
+
+        # Map aspect ratio / image_size to StepFun size string
+        size = _stepfun_size(aspect_ratio, image_size)
+        if size:
+            body["size"] = size
+
+        # step-1x-medium supports style_reference for reference-image-guided generation
+        refs = list(reference_images or [])
+        if refs and "1x" in model:
+            body["style_reference"] = {
+                "source_url": image_path_to_data_url(refs[0]),
+            }
+
+        body.update(self.extra_body)
+
+        response = await self._http_post(
+            f"{self.api_base}/images/generations",
+            headers=headers,
+            body=body,
+        )
+
+        try:
+            response.raise_for_status()
+        except httpx.HTTPStatusError as exc:
+            detail = response.text[:500]
+            raise ImageGenerationError(
+                f"StepFun image generation failed: {detail}"
+            ) from exc
+
+        payload = response.json()
+        images = _stepfun_images_from_payload(payload)
+
+        self._require_images(images, payload)
+
+        return GeneratedImageResponse(images=images, content="", raw=payload)
+
+
+def _stepfun_size(
+    aspect_ratio: str | None,
+    image_size: str | None,
+) -> str:
+    """Resolve aspect ratio / image_size to StepFun size string.
+
+    StepFun expects ``WIDTHxHEIGHT`` (note: width x height, not the more
+    common ``HxW`` order used by other providers).  The accepted sizes are
+    ``1024x1024``, ``768x1360``, ``896x1184``, ``1360x768``, ``1184x896``.
+    """
+    if image_size and "x" in image_size.lower():
+        return image_size
+    if aspect_ratio and aspect_ratio in _STEPFUN_ASPECT_RATIO_SIZES:
+        return _STEPFUN_ASPECT_RATIO_SIZES[aspect_ratio]
+    return "1024x1024"
+
+
+def _stepfun_images_from_payload(payload: dict[str, Any]) -> list[str]:
+    """Extract base64 images from StepFun API response.
+
+    StepFun returns images in ``data[].b64_json`` (base64 strings).
+    """
+    images: list[str] = []
+    for item in payload.get("data") or []:
+        if not isinstance(item, dict):
+            continue
+        b64 = item.get("b64_json")
+        if isinstance(b64, str) and b64:
+            images.append(_b64_image_data_url(b64))
+    return images
+
+
+# ---------------------------------------------------------------------------
+# Provider registration
+# ---------------------------------------------------------------------------
+
+register_image_gen_provider(OpenRouterImageGenerationClient)
+register_image_gen_provider(AIHubMixImageGenerationClient)
+register_image_gen_provider(GeminiImageGenerationClient)
+register_image_gen_provider(MiniMaxImageGenerationClient)
+register_image_gen_provider(StepFunImageGenerationClient)

nanobot/providers/openai_codex_provider.py

@@ -40,6 +40,7 @@ class OpenAICodexProvider(LLMProvider):
         reasoning_effort: str | None,
         tool_choice: str | dict[str, Any] | None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         """Shared request logic for both chat() and chat_stream()."""
         model = model or self.default_model
@@ -70,6 +71,7 @@ class OpenAICodexProvider(LLMProvider):
                 content, tool_calls, finish_reason = await _request_codex(
                     DEFAULT_CODEX_URL, headers, body, verify=True,
                     on_content_delta=on_content_delta,
+                    on_tool_call_delta=on_tool_call_delta,
                 )
             except Exception as e:
                 if "CERTIFICATE_VERIFY_FAILED" not in str(e):
@@ -78,6 +80,7 @@ class OpenAICodexProvider(LLMProvider):
                 content, tool_calls, finish_reason = await _request_codex(
                     DEFAULT_CODEX_URL, headers, body, verify=False,
                     on_content_delta=on_content_delta,
+                    on_tool_call_delta=on_tool_call_delta,
                 )
             return LLMResponse(content=content, tool_calls=tool_calls, finish_reason=finish_reason)
         except Exception as e:
@@ -100,9 +103,18 @@ class OpenAICodexProvider(LLMProvider):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         _ = on_thinking_delta
-        return await self._call_codex(messages, tools, model, reasoning_effort, tool_choice, on_content_delta)
+        return await self._call_codex(
+            messages,
+            tools,
+            model,
+            reasoning_effort,
+            tool_choice,
+            on_content_delta,
+            on_tool_call_delta,
+        )
 
     def get_default_model(self) -> str:
         return self.default_model
@@ -138,6 +150,7 @@ async def _request_codex(
     body: dict[str, Any],
     verify: bool,
     on_content_delta: Callable[[str], Awaitable[None]] | None = None,
+    on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
 ) -> tuple[str, list[ToolCallRequest], str]:
     async with httpx.AsyncClient(timeout=60.0, verify=verify) as client:
         async with client.stream("POST", url, headers=headers, json=body) as response:
@@ -148,7 +161,7 @@ async def _request_codex(
                     _friendly_error(response.status_code, text.decode("utf-8", "ignore")),
                     retry_after=retry_after,
                 )
-            return await consume_sse(response, on_content_delta)
+            return await consume_sse(response, on_content_delta, on_tool_call_delta)
 
 
 def _prompt_cache_key(messages: list[dict[str, Any]]) -> str:

nanobot/providers/openai_compat_provider.py

@@ -999,6 +999,21 @@ class OpenAICompatProvider(LLMProvider):
             if fn_prov:
                 buf["fn_prov"] = fn_prov
 
+        def _accum_legacy_function_call(function_call: Any) -> None:
+            """Accumulate legacy ``delta.function_call`` streaming chunks."""
+            if not function_call:
+                return
+            buf = tc_bufs.setdefault(0, {
+                "id": "", "name": "", "arguments": "",
+                "extra_content": None, "prov": None, "fn_prov": None,
+            })
+            fn_name = _get(function_call, "name")
+            if fn_name:
+                buf["name"] = str(fn_name)
+            fn_args = _get(function_call, "arguments")
+            if fn_args:
+                buf["arguments"] += str(fn_args)
+
         for chunk in chunks:
             if isinstance(chunk, str):
                 content_parts.append(chunk)
@@ -1029,6 +1044,7 @@ class OpenAICompatProvider(LLMProvider):
                     reasoning_parts.append(text)
                 for idx, tc in enumerate(delta.get("tool_calls") or []):
                     _accum_tc(tc, idx)
+                _accum_legacy_function_call(delta.get("function_call"))
                 usage = cls._extract_usage(chunk_map) or usage
                 continue
 
@@ -1047,8 +1063,10 @@ class OpenAICompatProvider(LLMProvider):
                     reasoning = getattr(delta, "reasoning", None)
                 if reasoning:
                     reasoning_parts.append(reasoning)
-            for tc in (delta.tool_calls or []) if delta else []:
+            for tc in (getattr(delta, "tool_calls", None) or []) if delta else []:
                 _accum_tc(tc, getattr(tc, "index", 0))
+            if delta:
+                _accum_legacy_function_call(getattr(delta, "function_call", None))
 
         return LLMResponse(
             content="".join(content_parts) or None,
@@ -1203,6 +1221,7 @@ class OpenAICompatProvider(LLMProvider):
         tool_choice: str | dict[str, Any] | None = None,
         on_content_delta: Callable[[str], Awaitable[None]] | None = None,
         on_thinking_delta: Callable[[str], Awaitable[None]] | None = None,
+        on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
     ) -> LLMResponse:
         idle_timeout_s = int(os.environ.get("NANOBOT_STREAM_IDLE_TIMEOUT_S", "90"))
         try:
@@ -1226,9 +1245,16 @@ class OpenAICompatProvider(LLMProvider):
                             except StopAsyncIteration:
                                 break
 
-                    content, tool_calls, finish_reason, usage, reasoning_content = await consume_sdk_stream(
+                    (
+                        content,
+                        tool_calls,
+                        finish_reason,
+                        usage,
+                        reasoning_content,
+                    ) = await consume_sdk_stream(
                         _timed_stream(),
                         on_content_delta,
+                        on_tool_call_delta=on_tool_call_delta,
                     )
                     self._record_responses_success(model, reasoning_effort)
                     return LLMResponse(
@@ -1252,6 +1278,12 @@ class OpenAICompatProvider(LLMProvider):
                 messages, tools, model, max_tokens, temperature,
                 reasoning_effort, tool_choice,
             )
+            if self._spec and self._spec.name == "zhipu" and tools and on_tool_call_delta:
+                # Z.AI/GLM keeps streaming tool-call arguments behind an
+                # explicit provider flag.  Pass it through the OpenAI SDK's
+                # extra_body escape hatch so the usual delta.tool_calls path
+                # can surface live file-edit progress.
+                kwargs.setdefault("extra_body", {})["tool_stream"] = True
             kwargs["stream"] = True
             kwargs["stream_options"] = {"include_usage": True}
             stream = await self._client.chat.completions.create(**kwargs)
@@ -1279,6 +1311,28 @@ class OpenAICompatProvider(LLMProvider):
                         r_text = self._extract_text_content(reasoning)
                         if r_text:
                             await on_thinking_delta(r_text)
+                    if on_tool_call_delta:
+                        for idx, tool_delta in enumerate(
+                            getattr(delta_obj, "tool_calls", None) or []
+                        ):
+                            fn = _get(tool_delta, "function")
+                            tool_index = _get(tool_delta, "index")
+                            await on_tool_call_delta({
+                                "index": tool_index if tool_index is not None else idx,
+                                "call_id": str(_get(tool_delta, "id") or ""),
+                                "name": str(_get(fn, "name") or "") if fn is not None else "",
+                                "arguments_delta": (
+                                    str(_get(fn, "arguments") or "") if fn is not None else ""
+                                ),
+                            })
+                        function_call = getattr(delta_obj, "function_call", None)
+                        if function_call:
+                            await on_tool_call_delta({
+                                "index": 0,
+                                "call_id": "",
+                                "name": str(_get(function_call, "name") or ""),
+                                "arguments_delta": str(_get(function_call, "arguments") or ""),
+                            })
             return self._parse_chunks(chunks)
         except asyncio.TimeoutError:
             return LLMResponse(

nanobot/providers/openai_responses/converters.py

@@ -5,6 +5,8 @@ from __future__ import annotations
 import json
 from typing import Any
 
+from nanobot.providers.base import LLMProvider
+
 
 def convert_messages(messages: list[dict[str, Any]]) -> tuple[str, list[dict[str, Any]]]:
     """Convert Chat Completions messages to Responses API input items.
@@ -58,8 +60,10 @@ def convert_messages(messages: list[dict[str, Any]]) -> tuple[str, list[dict[str
 def convert_user_message(content: Any) -> dict[str, Any]:
     """Convert a user message's content to Responses API format.
 
-    Handles plain strings, ``text`` blocks -> ``input_text``, and
-    ``image_url`` blocks -> ``input_image``.
+    Handles plain strings, ``text`` blocks -> ``input_text``,
+    ``image_url`` blocks -> ``input_image``, and ``input_audio`` blocks.
+    ``video_url`` is downgraded to a text placeholder because Codex does
+    not support native video.
     """
     if isinstance(content, str):
         return {"role": "user", "content": [{"type": "input_text", "text": content}]}
@@ -74,6 +78,18 @@ def convert_user_message(content: Any) -> dict[str, Any]:
                 url = (item.get("image_url") or {}).get("url")
                 if url:
                     converted.append({"type": "input_image", "image_url": url, "detail": "auto"})
+            elif item.get("type") == "input_audio":
+                audio_info = item.get("input_audio") or {}
+                audio_data = audio_info.get("data")
+                if audio_data:
+                    converted.append({
+                        "type": "input_audio",
+                        "input_audio": {"data": audio_data, "format": audio_info.get("format", "wav")},
+                    })
+            elif item.get("type") == "video_url":
+                # Codex doesn't support native video → text placeholder
+                placeholder = LLMProvider._media_placeholder("video_url", item)
+                converted.append({"type": "input_text", "text": placeholder["text"]})
         if converted:
             return {"role": "user", "content": converted}
     return {"role": "user", "content": [{"type": "input_text", "text": ""}]}

nanobot/providers/openai_responses/parsing.py

@@ -62,6 +62,7 @@ async def iter_sse(response: httpx.Response) -> AsyncGenerator[dict[str, Any], N
 async def consume_sse(
     response: httpx.Response,
     on_content_delta: Callable[[str], Awaitable[None]] | None = None,
+    on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
 ) -> tuple[str, list[ToolCallRequest], str]:
     """Consume a Responses API SSE stream into ``(content, tool_calls, finish_reason)``."""
     content = ""
@@ -82,6 +83,12 @@ async def consume_sse(
                     "name": item.get("name"),
                     "arguments": item.get("arguments") or "",
                 }
+                if on_tool_call_delta:
+                    await on_tool_call_delta({
+                        "call_id": str(call_id),
+                        "name": str(item.get("name") or ""),
+                        "arguments_delta": "",
+                    })
         elif event_type == "response.output_text.delta":
             delta_text = event.get("delta") or ""
             content += delta_text
@@ -90,7 +97,14 @@ async def consume_sse(
         elif event_type == "response.function_call_arguments.delta":
             call_id = event.get("call_id")
             if call_id and call_id in tool_call_buffers:
-                tool_call_buffers[call_id]["arguments"] += event.get("delta") or ""
+                delta = event.get("delta") or ""
+                tool_call_buffers[call_id]["arguments"] += delta
+                if on_tool_call_delta and delta:
+                    await on_tool_call_delta({
+                        "call_id": str(call_id),
+                        "name": str(tool_call_buffers[call_id].get("name") or ""),
+                        "arguments_delta": str(delta),
+                    })
         elif event_type == "response.function_call_arguments.done":
             call_id = event.get("call_id")
             if call_id and call_id in tool_call_buffers:
@@ -210,6 +224,7 @@ def parse_response_output(response: Any) -> LLMResponse:
 async def consume_sdk_stream(
     stream: Any,
     on_content_delta: Callable[[str], Awaitable[None]] | None = None,
+    on_tool_call_delta: Callable[[dict[str, Any]], Awaitable[None]] | None = None,
 ) -> tuple[str, list[ToolCallRequest], str, dict[str, int], str | None]:
     """Consume an SDK async stream from ``client.responses.create(stream=True)``."""
     content = ""
@@ -232,6 +247,12 @@ async def consume_sdk_stream(
                     "name": getattr(item, "name", None),
                     "arguments": getattr(item, "arguments", None) or "",
                 }
+                if on_tool_call_delta:
+                    await on_tool_call_delta({
+                        "call_id": str(call_id),
+                        "name": str(getattr(item, "name", None) or ""),
+                        "arguments_delta": "",
+                    })
         elif event_type == "response.output_text.delta":
             delta_text = getattr(event, "delta", "") or ""
             content += delta_text
@@ -240,7 +261,14 @@ async def consume_sdk_stream(
         elif event_type == "response.function_call_arguments.delta":
             call_id = getattr(event, "call_id", None)
             if call_id and call_id in tool_call_buffers:
-                tool_call_buffers[call_id]["arguments"] += getattr(event, "delta", "") or ""
+                delta = getattr(event, "delta", "") or ""
+                tool_call_buffers[call_id]["arguments"] += delta
+                if on_tool_call_delta and delta:
+                    await on_tool_call_delta({
+                        "call_id": str(call_id),
+                        "name": str(tool_call_buffers[call_id].get("name") or ""),
+                        "arguments_delta": str(delta),
+                    })
         elif event_type == "response.function_call_arguments.done":
             call_id = getattr(event, "call_id", None)
             if call_id and call_id in tool_call_buffers:

nanobot/providers/registry.py

@@ -155,6 +155,18 @@ PROVIDERS: tuple[ProviderSpec, ...] = (
         detect_by_base_keyword="huggingface",
         default_api_base="https://router.huggingface.co/v1",
     ),
+    # Skywork API platform (APIFree): OpenAI-compatible MaaS gateway.
+    ProviderSpec(
+        name="skywork",
+        keywords=("skywork", "skyclaw", "apifree"),
+        env_key="SKYWORK_API_KEY",
+        display_name="Skywork",
+        backend="openai_compat",
+        env_extras=(("APIFREE_API_KEY", "{api_key}"),),
+        is_gateway=True,
+        detect_by_base_keyword="apifree.ai",
+        default_api_base="https://api.apifree.ai/v1",
+    ),
     # AiHubMix: global gateway, OpenAI-compatible interface.
     # strip_model_prefix=True: doesn't understand "anthropic/claude-3",
     # strips to bare "claude-3".
@@ -390,13 +402,23 @@ PROVIDERS: tuple[ProviderSpec, ...] = (
         backend="openai_compat",
         default_api_base="https://api.longcat.chat/openai/v1",
     ),
+    # Ant Ling: OpenAI-compatible API for Ling/Ring model families.
+    ProviderSpec(
+        name="ant_ling",
+        keywords=("ant_ling", "ant-ling", "ling-", "ring-"),
+        env_key="ANT_LING_API_KEY",
+        display_name="Ant Ling",
+        backend="openai_compat",
+        detect_by_base_keyword="ant-ling.com",
+        default_api_base="https://api.ant-ling.com/v1",
+    ),
     # === Local deployment (matched by config key, NOT by api_base) =========
     # vLLM / any OpenAI-compatible local server
     ProviderSpec(
         name="vllm",
         keywords=("vllm",),
         env_key="HOSTED_VLLM_API_KEY",
-        display_name="vLLM/Local",
+        display_name="vLLM",
         backend="openai_compat",
         is_local=True,
     ),

nanobot/session/webui_turns.py

@@ -0,0 +1,347 @@
+"""Session turn helpers for WebUI-capable WebSocket sessions.
+
+AgentLoop uses these without importing a concrete channel plugin; only
+``channel == "websocket"`` messages are affected.
+"""
+
+from __future__ import annotations
+
+import re
+import time
+from collections.abc import Awaitable, Callable
+from dataclasses import dataclass, field
+from typing import Any
+
+from loguru import logger
+
+from nanobot.bus.events import InboundMessage, OutboundMessage
+from nanobot.bus.queue import MessageBus
+from nanobot.providers.base import LLMProvider
+from nanobot.session.goal_state import goal_state_ws_blob
+from nanobot.session.manager import Session, SessionManager
+from nanobot.utils.helpers import truncate_text
+from nanobot.utils.llm_runtime import LLMRuntime
+
+WEBUI_SESSION_METADATA_KEY = "webui"
+WEBUI_TITLE_METADATA_KEY = "title"
+WEBUI_TITLE_USER_EDITED_METADATA_KEY = "title_user_edited"
+TITLE_MAX_CHARS = 60
+TITLE_GENERATION_MAX_TOKENS = 96
+TITLE_GENERATION_REASONING_EFFORT = "none"
+
+# Wall-clock turn start per ``chat_id`` (websocket only). Survives browser refresh while the
+# gateway process stays up; cleared on idle/stop and implicitly dropped on restart.
+_WEBSOCKET_TURN_WALL_STARTED_AT: dict[str, float] = {}
+
+
+def mark_webui_session(session: Session, metadata: dict[str, Any]) -> bool:
+    """Persist a WebUI marker only when the inbound websocket frame opted in."""
+    if metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
+        return False
+    session.metadata[WEBUI_SESSION_METADATA_KEY] = True
+    return True
+
+
+def clean_generated_title(raw: str | None) -> str:
+    text = (raw or "").strip()
+    if not text:
+        return ""
+    text = re.sub(r"^\s*(title|标题)\s*[:：]\s*", "", text, flags=re.IGNORECASE)
+    text = text.strip().strip("\"'`“”‘’")
+    text = re.sub(r"\s+", " ", text).strip()
+    text = text.rstrip("。.!！?？,，;；:")
+    if len(text) > TITLE_MAX_CHARS:
+        text = text[: TITLE_MAX_CHARS - 1].rstrip() + "…"
+    return text
+
+
+def _title_inputs(session: Session) -> tuple[str, str]:
+    user_text = ""
+    assistant_text = ""
+    for message in session.messages:
+        if message.get("_command") is True:
+            continue
+        role = message.get("role")
+        content = message.get("content")
+        if not isinstance(content, str) or not content.strip():
+            continue
+        if role == "user" and not user_text:
+            user_text = content.strip()
+        elif role == "assistant" and not assistant_text:
+            assistant_text = content.strip()
+        if user_text and assistant_text:
+            break
+    return user_text, assistant_text
+
+
+async def maybe_generate_webui_title(
+    *,
+    sessions: SessionManager,
+    session_key: str,
+    provider: LLMProvider,
+    model: str,
+) -> bool:
+    """Generate and persist a short title for WebUI-owned sessions only."""
+    session = sessions.get_or_create(session_key)
+    if session.metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
+        return False
+    if session.metadata.get(WEBUI_TITLE_USER_EDITED_METADATA_KEY) is True:
+        return False
+    current_title = session.metadata.get(WEBUI_TITLE_METADATA_KEY)
+    if isinstance(current_title, str) and current_title.strip():
+        return False
+
+    user_text, assistant_text = _title_inputs(session)
+    if not user_text:
+        return False
+
+    prompt = (
+        "Generate a concise title for this chat.\n"
+        "Rules:\n"
+        "- Use the same language as the user when practical.\n"
+        "- 3 to 8 words.\n"
+        "- No quotes.\n"
+        "- No punctuation at the end.\n"
+        "- Return only the title.\n\n"
+        f"User: {truncate_text(user_text, 1_000)}"
+    )
+    if assistant_text:
+        prompt += f"\nAssistant: {truncate_text(assistant_text, 1_000)}"
+
+    try:
+        response = await provider.chat_with_retry(
+            [
+                {
+                    "role": "system",
+                    "content": (
+                        "You write short, neutral chat titles. "
+                        "Return only the title text."
+                    ),
+                },
+                {"role": "user", "content": prompt},
+            ],
+            tools=None,
+            model=model,
+            max_tokens=TITLE_GENERATION_MAX_TOKENS,
+            temperature=0.2,
+            reasoning_effort=TITLE_GENERATION_REASONING_EFFORT,
+            retry_mode="standard",
+        )
+    except Exception:
+        logger.debug("Failed to generate webui session title for {}", session_key, exc_info=True)
+        return False
+
+    title = clean_generated_title(response.content)
+    if not title or title.lower().startswith("error"):
+        logger.debug(
+            "WebUI title generation returned no usable title for {} (finish_reason={})",
+            session_key,
+            response.finish_reason,
+        )
+        return False
+    session.metadata[WEBUI_TITLE_METADATA_KEY] = title
+    sessions.save(session)
+    return True
+
+
+async def maybe_generate_webui_title_after_turn(
+    *,
+    channel: str,
+    metadata: dict[str, Any],
+    sessions: SessionManager,
+    session_key: str,
+    provider: LLMProvider,
+    model: str,
+) -> bool:
+    if channel != "websocket" or metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
+        return False
+    return await maybe_generate_webui_title(
+        sessions=sessions,
+        session_key=session_key,
+        provider=provider,
+        model=model,
+    )
+
+
+def websocket_turn_wall_started_at(chat_id: str) -> float | None:
+    """Return ``time.time()`` when the active user turn began, if still running."""
+    return _WEBSOCKET_TURN_WALL_STARTED_AT.get(chat_id)
+
+
+async def publish_turn_run_status(bus: MessageBus, msg: InboundMessage, status: str) -> None:
+    """Notify WebSocket clients while a user turn is executing (timing strip)."""
+    if msg.channel != "websocket":
+        return
+    cid = str(msg.chat_id)
+    meta: dict[str, Any] = {
+        **dict(msg.metadata or {}),
+        "_goal_status": True,
+        "goal_status": status,
+    }
+    if status == "running":
+        t0 = time.time()
+        meta["started_at"] = t0
+        _WEBSOCKET_TURN_WALL_STARTED_AT[cid] = t0
+    else:
+        _WEBSOCKET_TURN_WALL_STARTED_AT.pop(cid, None)
+    await bus.publish_outbound(
+        OutboundMessage(
+            channel=msg.channel,
+            chat_id=cid,
+            content="",
+            metadata=meta,
+        ),
+    )
+
+
+def build_bus_progress_callback(
+    bus: MessageBus,
+    msg: InboundMessage,
+) -> Callable[..., Awaitable[None]]:
+    """Return the bus progress callback for agent runtime events."""
+
+    async def _publish_progress(
+        content: str,
+        *,
+        tool_hint: bool = False,
+        tool_events: list[dict[str, Any]] | None = None,
+        file_edit_events: list[dict[str, Any]] | None = None,
+        reasoning: bool = False,
+        reasoning_end: bool = False,
+    ) -> None:
+        meta = dict(msg.metadata or {})
+        meta["_progress"] = True
+        meta["_tool_hint"] = tool_hint
+        if reasoning:
+            meta["_reasoning_delta"] = True
+        if reasoning_end:
+            meta["_reasoning_end"] = True
+        if tool_events:
+            meta["_tool_events"] = tool_events
+        if file_edit_events:
+            meta["_file_edit_events"] = file_edit_events
+        await bus.publish_outbound(
+            OutboundMessage(
+                channel=msg.channel,
+                chat_id=msg.chat_id,
+                content=content,
+                metadata=meta,
+            )
+        )
+
+    if msg.channel == "websocket":
+        async def _websocket_progress(
+            content: str,
+            *,
+            tool_hint: bool = False,
+            tool_events: list[dict[str, Any]] | None = None,
+            file_edit_events: list[dict[str, Any]] | None = None,
+            reasoning: bool = False,
+            reasoning_end: bool = False,
+        ) -> None:
+            await _publish_progress(
+                content,
+                tool_hint=tool_hint,
+                tool_events=tool_events,
+                file_edit_events=file_edit_events,
+                reasoning=reasoning,
+                reasoning_end=reasoning_end,
+            )
+
+        return _websocket_progress
+
+    async def _bus_progress(
+        content: str,
+        *,
+        tool_hint: bool = False,
+        tool_events: list[dict[str, Any]] | None = None,
+        reasoning: bool = False,
+        reasoning_end: bool = False,
+    ) -> None:
+        await _publish_progress(
+            content,
+            tool_hint=tool_hint,
+            tool_events=tool_events,
+            reasoning=reasoning,
+            reasoning_end=reasoning_end,
+        )
+
+    return _bus_progress
+
+
+@dataclass
+class WebuiTurnCoordinator:
+    """Own the WebUI/WebSocket wire details that hang off AgentLoop turns."""
+
+    bus: MessageBus
+    sessions: SessionManager
+    schedule_background: Callable[[Awaitable[None]], None]
+    _title_contexts: dict[str, LLMRuntime] = field(default_factory=dict)
+
+    def capture_title_context(
+        self,
+        session_key: str,
+        msg: InboundMessage,
+        llm: LLMRuntime,
+    ) -> None:
+        if msg.channel == "websocket" and msg.metadata.get("webui") is True:
+            self._title_contexts[session_key] = llm
+
+    def discard(self, session_key: str) -> None:
+        self._title_contexts.pop(session_key, None)
+
+    async def publish_run_status(self, msg: InboundMessage, status: str) -> None:
+        await publish_turn_run_status(self.bus, msg, status)
+
+    async def handle_turn_end(
+        self,
+        msg: InboundMessage,
+        *,
+        session_key: str,
+        latency_ms: int | None,
+    ) -> None:
+        if msg.channel != "websocket":
+            return
+
+        turn_metadata: dict[str, Any] = {**msg.metadata, "_turn_end": True}
+        if latency_ms is not None:
+            turn_metadata["latency_ms"] = int(latency_ms)
+        session = self.sessions.get_or_create(session_key)
+        turn_metadata["goal_state"] = goal_state_ws_blob(session.metadata)
+        await self.bus.publish_outbound(OutboundMessage(
+            channel=msg.channel,
+            chat_id=msg.chat_id,
+            content="",
+            metadata=turn_metadata,
+        ))
+        self._schedule_title_update(msg, session_key=session_key)
+
+    def _schedule_title_update(self, msg: InboundMessage, *, session_key: str) -> None:
+        title_context = self._title_contexts.pop(session_key, None)
+        if msg.metadata.get("webui") is not True or title_context is None:
+            return
+
+        async def _generate_title_and_notify(
+            title_llm: LLMRuntime = title_context,
+        ) -> None:
+            generated = await maybe_generate_webui_title_after_turn(
+                channel=msg.channel,
+                metadata=msg.metadata,
+                sessions=self.sessions,
+                session_key=session_key,
+                provider=title_llm.provider,
+                model=title_llm.model,
+            )
+            if generated:
+                await self.bus.publish_outbound(OutboundMessage(
+                    channel=msg.channel,
+                    chat_id=msg.chat_id,
+                    content="",
+                    metadata={
+                        **msg.metadata,
+                        "_session_updated": True,
+                        "_session_update_scope": "metadata",
+                    },
+                ))
+
+        self.schedule_background(_generate_title_and_notify())

nanobot/skills/image-generation/SKILL.md

@@ -15,7 +15,7 @@ If the `generate_image` tool is not available in the current tool list, tell the
 - Image editing: pass the saved artifact path or user image path in `reference_images`.
 - Iterative edits in the same conversation: prefer the most recent generated image artifact if the user says things like "make it brighter", "change the background", or "try another version".
 - Ambiguous edits: ask a short clarifying question if multiple recent images could be the target.
-- In the current chat, do not call `message` just to announce or resend generated images. The runtime attaches images from `generate_image` to the final assistant reply automatically.
+- After generating images, call the `message` tool with the artifact paths in the `media` parameter to deliver them to the user.
 
 ## Prompt Rules
 
@@ -42,52 +42,6 @@ For follow-up edits, pass the prior artifact `path` to `reference_images`. If th
 
 Do not include internal replay markers such as `[Message Time: ...]`, `[image: /local/path]`, `generate_image(...)`, or `message(...)` in user-facing replies.
 
-## Provider Notes
-
-Do not ask users to paste API keys into chat. If configuration is needed, describe the fields; LLM provider and BYOK changes are hot-reloaded for new turns.
-
-For OpenRouter, the image tool expects:
-
-```json
-{
-  "providers": {
-    "openrouter": {
-      "apiKey": "sk-or-..."
-    }
-  },
-  "tools": {
-    "imageGeneration": {
-      "enabled": true,
-      "provider": "openrouter",
-      "model": "openai/gpt-5.4-image-2"
-    }
-  }
-}
-```
-
-For AIHubMix, the image tool expects:
-
-```json
-{
-  "providers": {
-    "aihubmix": {
-      "apiKey": "sk-..."
-    }
-  },
-  "tools": {
-    "imageGeneration": {
-      "enabled": true,
-      "provider": "aihubmix",
-      "model": "gpt-image-2-free"
-    }
-  }
-}
-```
-
-AIHubMix `gpt-image-2-free` uses AIHubMix's unified predictions endpoint internally (`/v1/models/openai/gpt-image-2-free/predictions`), not the OpenAI Images `/v1/images/generations` endpoint. If it fails with "Incorrect model ID", do not assume the key lacks permission until the provider config, model name, and gateway restart have been checked.
-
-`providers.aihubmix.extraBody` can be used for provider-specific options. For example, `"extraBody": {"quality": "low"}` is optional but can make `gpt-image-2-free` faster and less likely to time out.
-
 ## Examples
 
 Generate a new image:

nanobot/templates/agent/identity.md

@@ -30,5 +30,5 @@ Output is rendered in a terminal. Avoid markdown headings and tables. Use plain
 
 Reply directly with text for the current conversation. Do not use the 'message' tool for normal replies in the current chat.
 When you need to call tools before answering, do not include the final user-visible answer in the same assistant message as the tool calls. Wait for the tool results, then answer once.
-Use the 'message' tool only for proactive sends, cross-channel delivery, or explicitly sending existing local files as attachments. When a tool such as 'generate_image' creates user-visible media, the runtime attaches those artifacts to the final assistant reply automatically, so do not call 'message' just to announce or resend them.
+Use the 'message' tool only for proactive sends, cross-channel delivery, or explicitly sending existing local files as attachments. When 'generate_image' creates images, call 'message' with the artifact paths in the 'media' parameter to deliver them to the user.
 To send an existing local file that was not automatically attached by another tool, call 'message' with the 'media' parameter. Do NOT use read_file to "send" a file — reading a file only shows its content to you, it does NOT deliver the file to the user. Example: message(content="Here is the document", channel="telegram", chat_id="...", media=["/path/to/file.pdf"])

nanobot/utils/__init__.py

@@ -1,6 +1,42 @@
 """Utility functions for nanobot."""
 
+from __future__ import annotations
+
+import sys
+from importlib import import_module
+from types import ModuleType
+
 from nanobot.utils.helpers import ensure_dir
 from nanobot.utils.path import abbreviate_path
 
 __all__ = ["ensure_dir", "abbreviate_path"]
+
+
+class _LazyModuleAlias(ModuleType):
+    def __init__(self, name: str, target: str) -> None:
+        super().__init__(name)
+        self.__dict__["_target"] = target
+
+    def _load(self) -> ModuleType:
+        module = import_module(self.__dict__["_target"])
+        sys.modules[self.__name__] = module
+        return module
+
+    def __getattr__(self, name: str) -> object:
+        return getattr(self._load(), name)
+
+    def __dir__(self) -> list[str]:
+        return sorted(set(super().__dir__()) | set(dir(self._load())))
+
+
+_LEGACY_MODULE_ALIASES = {
+    "webui_thread_disk": "nanobot.webui.thread_disk",
+    "webui_transcript": "nanobot.webui.transcript",
+    "webui_turn_helpers": "nanobot.session.webui_turns",
+}
+
+for _legacy_name, _target_name in _LEGACY_MODULE_ALIASES.items():
+    sys.modules.setdefault(
+        f"{__name__}.{_legacy_name}",
+        _LazyModuleAlias(f"{__name__}.{_legacy_name}", _target_name),
+    )

nanobot/utils/artifacts.py

@@ -21,8 +21,6 @@ _MIME_EXTENSIONS = {
     "image/webp": ".webp",
     "image/gif": ".gif",
 }
-_GENERATE_IMAGE_TOOL_NAME = "generate_image"
-
 
 class ArtifactError(ValueError):
     """Raised when an artifact cannot be safely decoded or stored."""
@@ -115,48 +113,10 @@ def generated_image_tool_result(artifacts: list[dict[str, Any]]) -> str:
             "artifacts": artifacts,
             "next_step": (
                 "Use these artifact paths as reference_images for follow-up edits. "
-                "For the current chat, reply naturally; the runtime attaches generated images automatically. "
-                "Do not call message just to announce or resend them. Keep raw paths internal unless the user asks for debug details."
+                "Call the message tool with the artifact paths in the media parameter "
+                "to deliver the images to the user. Keep raw paths internal unless the "
+                "user asks for debug details."
             ),
         },
         ensure_ascii=False,
     )
-
-
-def _extract_text_payload(content: Any) -> str | None:
-    if isinstance(content, str):
-        return content
-    if isinstance(content, list):
-        parts: list[str] = []
-        for block in content:
-            if isinstance(block, dict) and isinstance(block.get("text"), str):
-                parts.append(block["text"])
-        return "\n".join(parts) if parts else None
-    return None
-
-
-def generated_image_paths_from_messages(messages: list[dict[str, Any]]) -> list[str]:
-    """Collect generated image artifact paths from generate_image tool results."""
-    paths: list[str] = []
-    seen: set[str] = set()
-    for message in messages:
-        if message.get("role") != "tool" or message.get("name") != _GENERATE_IMAGE_TOOL_NAME:
-            continue
-        payload = _extract_text_payload(message.get("content"))
-        if not payload:
-            continue
-        try:
-            data = json.loads(payload)
-        except json.JSONDecodeError:
-            continue
-        artifacts = data.get("artifacts") if isinstance(data, dict) else None
-        if not isinstance(artifacts, list):
-            continue
-        for artifact in artifacts:
-            if not isinstance(artifact, dict):
-                continue
-            path = artifact.get("path")
-            if isinstance(path, str) and path and path not in seen:
-                paths.append(path)
-                seen.add(path)
-    return paths

nanobot/utils/file_edit_events.py

@@ -0,0 +1,780 @@
+"""File-edit activity helpers for WebUI progress events."""
+
+from __future__ import annotations
+
+import difflib
+import json
+import re
+import time
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Awaitable, Callable
+
+
+TRACKED_FILE_EDIT_TOOLS = frozenset({"write_file", "edit_file", "notebook_edit"})
+_MAX_SNAPSHOT_BYTES = 2 * 1024 * 1024
+_LIVE_EMIT_INTERVAL_S = 0.18
+_LIVE_EMIT_LINE_STEP = 24
+
+
+@dataclass(slots=True)
+class FileSnapshot:
+    path: Path
+    exists: bool
+    text: str | None
+    unreadable: bool = False
+    binary: bool = False
+    oversized: bool = False
+
+    @property
+    def countable(self) -> bool:
+        return (
+            self.text is not None
+            and not self.binary
+            and not self.oversized
+            and not self.unreadable
+        )
+
+
+@dataclass(slots=True)
+class FileEditTracker:
+    call_id: str
+    tool: str
+    path: Path
+    display_path: str
+    before: FileSnapshot
+
+
+def is_file_edit_tool(tool_name: str | None) -> bool:
+    return bool(tool_name) and tool_name in TRACKED_FILE_EDIT_TOOLS
+
+
+def resolve_file_edit_path(
+    tool: Any,
+    workspace: Path | None,
+    params: dict[str, Any] | None,
+) -> Path | None:
+    """Resolve the target file path after tool argument preparation."""
+    if not isinstance(params, dict):
+        return None
+    raw_path = params.get("path")
+    if not isinstance(raw_path, str) or not raw_path.strip():
+        return None
+    resolver = getattr(tool, "_resolve", None)
+    if callable(resolver):
+        try:
+            resolved = resolver(raw_path)
+            if isinstance(resolved, Path):
+                return resolved
+            if resolved:
+                return Path(resolved)
+        except Exception:
+            return None
+    if workspace is None:
+        return Path(raw_path).expanduser().resolve()
+    return (workspace / raw_path).expanduser().resolve()
+
+
+def display_file_edit_path(path: Path, workspace: Path | None) -> str:
+    if workspace is not None:
+        try:
+            return path.resolve().relative_to(workspace.resolve()).as_posix()
+        except Exception:
+            pass
+    return path.as_posix()
+
+
+def read_file_snapshot(path: Path, *, max_bytes: int = _MAX_SNAPSHOT_BYTES) -> FileSnapshot:
+    try:
+        if not path.exists() or not path.is_file():
+            return FileSnapshot(path=path, exists=False, text="")
+        size = path.stat().st_size
+        if size > max_bytes:
+            return FileSnapshot(path=path, exists=True, text=None, oversized=True)
+        raw = path.read_bytes()
+    except OSError:
+        return FileSnapshot(path=path, exists=path.exists(), text=None, unreadable=True)
+    if b"\x00" in raw:
+        return FileSnapshot(path=path, exists=True, text=None, binary=True)
+    try:
+        text = raw.decode("utf-8")
+    except UnicodeDecodeError:
+        return FileSnapshot(path=path, exists=True, text=None, binary=True)
+    return FileSnapshot(path=path, exists=True, text=text.replace("\r\n", "\n"))
+
+
+def line_diff_stats(before: str | None, after: str | None) -> tuple[int, int]:
+    """Return ``(added, deleted)`` for a UTF-8 text line-level diff."""
+    if before is None or after is None:
+        return 0, 0
+    if before == "":
+        return _text_line_count(after), 0
+    before_lines = before.replace("\r\n", "\n").splitlines()
+    after_lines = after.replace("\r\n", "\n").splitlines()
+    added = 0
+    deleted = 0
+    matcher = difflib.SequenceMatcher(a=before_lines, b=after_lines, autojunk=False)
+    for tag, i1, i2, j1, j2 in matcher.get_opcodes():
+        if tag == "equal":
+            continue
+        if tag in ("replace", "delete"):
+            deleted += i2 - i1
+        if tag in ("replace", "insert"):
+            added += j2 - j1
+    return added, deleted
+
+
+def _text_line_count(text: str) -> int:
+    if not text:
+        return 0
+    line_count = 0
+    last_was_newline = False
+    last_was_cr = False
+    for ch in text:
+        if ch == "\r":
+            line_count += 1
+            last_was_newline = True
+            last_was_cr = True
+        elif ch == "\n":
+            if not last_was_cr:
+                line_count += 1
+            last_was_newline = True
+            last_was_cr = False
+        else:
+            last_was_newline = False
+            last_was_cr = False
+    return line_count if last_was_newline else line_count + 1
+
+
+def prepare_file_edit_tracker(
+    *,
+    call_id: str,
+    tool_name: str,
+    tool: Any,
+    workspace: Path | None,
+    params: dict[str, Any] | None,
+) -> FileEditTracker | None:
+    if not is_file_edit_tool(tool_name):
+        return None
+    path = resolve_file_edit_path(tool, workspace, params)
+    if path is None:
+        return None
+    before = read_file_snapshot(path)
+    return FileEditTracker(
+        call_id=str(call_id or ""),
+        tool=tool_name,
+        path=path,
+        display_path=display_file_edit_path(path, workspace),
+        before=before,
+    )
+
+
+def build_file_edit_start_event(
+    tracker: FileEditTracker,
+    params: dict[str, Any] | None,
+) -> dict[str, Any]:
+    predicted_after = _predict_after_text(tracker.tool, params or {}, tracker.before)
+    if tracker.before.countable and predicted_after is not None:
+        added, deleted = line_diff_stats(tracker.before.text, predicted_after)
+    else:
+        added, deleted = 0, 0
+    return _event_payload(
+        tracker,
+        phase="start",
+        status="editing",
+        added=added,
+        deleted=deleted,
+        approximate=True,
+    )
+
+
+def build_file_edit_end_event(
+    tracker: FileEditTracker,
+    params: dict[str, Any] | None = None,
+) -> dict[str, Any]:
+    after = read_file_snapshot(tracker.path)
+    counted = False
+    if tracker.before.countable and after.countable:
+        added, deleted = line_diff_stats(tracker.before.text, after.text)
+        counted = True
+    else:
+        predicted_after = _predict_after_text(tracker.tool, params or {}, tracker.before)
+        if tracker.before.countable and predicted_after is not None:
+            added, deleted = line_diff_stats(tracker.before.text, predicted_after)
+            counted = True
+        else:
+            added, deleted = 0, 0
+    return _event_payload(
+        tracker,
+        phase="end",
+        status="done",
+        added=added,
+        deleted=deleted,
+        approximate=False,
+        binary=(after.binary or after.oversized or after.unreadable) and not counted,
+    )
+
+
+def build_file_edit_error_event(
+    tracker: FileEditTracker,
+    error: str | None = None,
+) -> dict[str, Any]:
+    payload = _event_payload(
+        tracker,
+        phase="error",
+        status="error",
+        added=0,
+        deleted=0,
+        approximate=False,
+    )
+    if error:
+        payload["error"] = error.strip()[:240]
+    return payload
+
+
+def build_file_edit_live_event(
+    tracker: FileEditTracker,
+    *,
+    added: int,
+    deleted: int = 0,
+) -> dict[str, Any]:
+    """Build an approximate in-progress event while tool-call arguments stream."""
+    return _event_payload(
+        tracker,
+        phase="start",
+        status="editing",
+        added=added,
+        deleted=deleted,
+        approximate=True,
+    )
+
+
+def build_file_edit_pending_event(
+    *,
+    call_id: str,
+    tool_name: str,
+    added: int = 0,
+    deleted: int = 0,
+) -> dict[str, Any]:
+    """Build an early placeholder before the streamed JSON path is available."""
+    return {
+        "version": 1,
+        "call_id": str(call_id or ""),
+        "tool": tool_name,
+        "path": "",
+        "phase": "start",
+        "added": max(0, int(added)),
+        "deleted": max(0, int(deleted)),
+        "approximate": True,
+        "status": "editing",
+        "pending": True,
+    }
+
+
+class StreamingFileEditTracker:
+    """Track file-edit tool arguments while the model is still streaming them.
+
+    Tool execution events only begin after the provider has completed the full
+    function call.  For large ``write_file`` calls, the long wait is usually the
+    model producing the JSON ``content`` argument.  Large ``edit_file`` calls
+    can have the same wait while ``old_text`` / ``new_text`` stream in.  This
+    tracker converts those argument deltas into approximate WebUI file-edit
+    events before the final exact diff is available.
+    """
+
+    def __init__(
+        self,
+        *,
+        workspace: Path | None,
+        tools: Any,
+        emit: Callable[[list[dict[str, Any]]], Awaitable[None]],
+    ) -> None:
+        self._workspace = workspace
+        self._tools = tools
+        self._emit = emit
+        self._states: dict[str, _StreamingFileEditState] = {}
+
+    async def update(self, payload: dict[str, Any]) -> None:
+        key = _stream_key(payload)
+        if not key:
+            return
+        state = self._states.get(key)
+        if state is None:
+            state = _StreamingFileEditState(key=key)
+            self._states[key] = state
+
+        state.apply_delta(payload)
+        if state.name not in {"write_file", "edit_file"}:
+            return
+        if state.path is None:
+            state.path = _extract_complete_json_string(state.arguments, "path")
+        if state.path is None:
+            added, deleted = state.live_diff_counts()
+            now = time.monotonic()
+            if state.should_emit_pending(added, deleted, now):
+                state.mark_pending_emitted(added, deleted, now)
+                await self._emit([build_file_edit_pending_event(
+                    call_id=state.call_id or state.key,
+                    tool_name=state.name,
+                    added=added,
+                    deleted=deleted,
+                )])
+            return
+        if state.tracker is None:
+            tool = self._tools.get(state.name) if hasattr(self._tools, "get") else None
+            state.tracker = prepare_file_edit_tracker(
+                call_id=state.call_id or state.key,
+                tool_name=state.name,
+                tool=tool,
+                workspace=self._workspace,
+                params={"path": state.path},
+            )
+            if state.tracker is None:
+                return
+
+        added, deleted = state.live_diff_counts()
+        now = time.monotonic()
+        if not state.should_emit(added, deleted, now):
+            return
+        state.mark_emitted(added, deleted, now)
+        await self._emit([build_file_edit_live_event(
+            state.tracker,
+            added=added,
+            deleted=deleted,
+        )])
+
+    async def flush(self) -> None:
+        events: list[dict[str, Any]] = []
+        now = time.monotonic()
+        for state in self._states.values():
+            if state.tracker is None:
+                continue
+            added, deleted = state.live_diff_counts()
+            if (
+                state.last_emitted_added == added
+                and state.last_emitted_deleted == deleted
+                and state.emitted_once
+            ):
+                continue
+            state.mark_emitted(added, deleted, now)
+            events.append(build_file_edit_live_event(
+                state.tracker,
+                added=added,
+                deleted=deleted,
+            ))
+        if events:
+            await self._emit(events)
+
+    def apply_final_call_ids(self, final_tool_calls: list[Any]) -> None:
+        """Keep final start/end events keyed to any earlier streamed placeholder."""
+        for tool_call in final_tool_calls:
+            canonical = self.canonical_call_id_for(tool_call)
+            if canonical:
+                try:
+                    tool_call.id = canonical
+                except Exception:
+                    pass
+
+    def canonical_call_id_for(self, tool_call: Any) -> str | None:
+        for state in self._states.values():
+            if state.matches_final_tool_call(tool_call):
+                return state.call_id or (state.tracker.call_id if state.tracker else None) or state.key
+        return None
+
+    async def error_unmatched(
+        self,
+        final_tool_calls: list[Any],
+        error: str,
+    ) -> None:
+        """Mark streamed edits as failed when no final tool call will run."""
+        events: list[dict[str, Any]] = []
+        for state in self._states.values():
+            if state.tracker is None:
+                continue
+            if any(state.matches_final_tool_call(tool_call) for tool_call in final_tool_calls):
+                continue
+            events.append(build_file_edit_error_event(state.tracker, error))
+        if events:
+            await self._emit(events)
+
+
+@dataclass(slots=True)
+class _StreamingJsonStringField:
+    key: str
+    scan_pos: int | None = None
+    closed: bool = False
+    escape: bool = False
+    unicode_remaining: int = 0
+    unicode_buffer: str = ""
+    newline_count: int = 0
+    has_chars: bool = False
+    last_char_newline: bool = False
+    last_char_cr: bool = False
+
+    @property
+    def line_count(self) -> int:
+        if not self.has_chars:
+            return 0
+        return self.newline_count + (0 if self.last_char_newline else 1)
+
+    def reset(self) -> None:
+        self.scan_pos = None
+        self.closed = False
+        self.escape = False
+        self.unicode_remaining = 0
+        self.unicode_buffer = ""
+        self.newline_count = 0
+        self.has_chars = False
+        self.last_char_newline = False
+        self.last_char_cr = False
+
+    def scan(self, source: str) -> None:
+        if self.closed:
+            return
+        if self.scan_pos is None:
+            match = re.search(rf'"{re.escape(self.key)}"\s*:\s*"', source)
+            if match is None:
+                return
+            self.scan_pos = match.end()
+        i = self.scan_pos
+        while i < len(source):
+            ch = source[i]
+            if self.unicode_remaining > 0:
+                self.unicode_buffer += ch
+                self.unicode_remaining -= 1
+                if self.unicode_remaining == 0:
+                    try:
+                        decoded = chr(int(self.unicode_buffer, 16))
+                    except ValueError:
+                        decoded = "x"
+                    self.unicode_buffer = ""
+                    self._mark_char(decoded)
+                i += 1
+                continue
+            if self.escape:
+                self.escape = False
+                if ch == "u":
+                    self.unicode_remaining = 4
+                    self.unicode_buffer = ""
+                elif ch == "n":
+                    self._mark_char("\n")
+                elif ch == "r":
+                    self._mark_char("\r")
+                else:
+                    self._mark_char(ch)
+                i += 1
+                continue
+            if ch == "\\":
+                self.escape = True
+                i += 1
+                continue
+            if ch == '"':
+                self.closed = True
+                i += 1
+                break
+            self._mark_char(ch)
+            i += 1
+        self.scan_pos = i
+
+    def _mark_char(self, ch: str) -> None:
+        self.has_chars = True
+        if ch == "\r":
+            self.newline_count += 1
+            self.last_char_newline = True
+            self.last_char_cr = True
+        elif ch == "\n":
+            if not self.last_char_cr:
+                self.newline_count += 1
+            self.last_char_newline = True
+            self.last_char_cr = False
+        else:
+            self.last_char_newline = False
+            self.last_char_cr = False
+
+
+@dataclass(slots=True)
+class _StreamingFileEditState:
+    key: str
+    call_id: str = ""
+    name: str = ""
+    arguments: str = ""
+    path: str | None = None
+    tracker: FileEditTracker | None = None
+    content: _StreamingJsonStringField = field(
+        default_factory=lambda: _StreamingJsonStringField("content")
+    )
+    old_text: _StreamingJsonStringField = field(
+        default_factory=lambda: _StreamingJsonStringField("old_text")
+    )
+    new_text: _StreamingJsonStringField = field(
+        default_factory=lambda: _StreamingJsonStringField("new_text")
+    )
+    emitted_once: bool = False
+    last_emitted_added: int = -1
+    last_emitted_deleted: int = -1
+    last_emit_at: float = 0.0
+    pending_emitted: bool = False
+    last_pending_added: int = -1
+    last_pending_deleted: int = -1
+    last_pending_at: float = 0.0
+
+    def apply_delta(self, payload: dict[str, Any]) -> None:
+        call_id = payload.get("call_id")
+        if isinstance(call_id, str) and call_id:
+            self.call_id = call_id
+        name = payload.get("name")
+        if isinstance(name, str) and name:
+            self.name = name
+        args = payload.get("arguments")
+        if isinstance(args, str):
+            self.arguments = args
+            self.content.reset()
+            self.old_text.reset()
+            self.new_text.reset()
+            return
+        delta = payload.get("arguments_delta")
+        if isinstance(delta, str) and delta:
+            self.arguments += delta
+
+    def live_diff_counts(self) -> tuple[int, int]:
+        if self.name == "write_file":
+            self.content.scan(self.arguments)
+            return self.content.line_count, 0
+        if self.name == "edit_file":
+            self.old_text.scan(self.arguments)
+            self.new_text.scan(self.arguments)
+            return self.new_text.line_count, self.old_text.line_count
+        return 0, 0
+
+    def should_emit(self, added: int, deleted: int, now: float) -> bool:
+        if not self.emitted_once:
+            return True
+        if added == self.last_emitted_added and deleted == self.last_emitted_deleted:
+            return False
+        if max(
+            abs(added - self.last_emitted_added),
+            abs(deleted - self.last_emitted_deleted),
+        ) >= _LIVE_EMIT_LINE_STEP:
+            return True
+        return now - self.last_emit_at >= _LIVE_EMIT_INTERVAL_S
+
+    def mark_emitted(self, added: int, deleted: int, now: float) -> None:
+        self.emitted_once = True
+        self.last_emitted_added = added
+        self.last_emitted_deleted = deleted
+        self.last_emit_at = now
+
+    def should_emit_pending(self, added: int, deleted: int, now: float) -> bool:
+        if not self.pending_emitted:
+            return True
+        if added == self.last_pending_added and deleted == self.last_pending_deleted:
+            return False
+        if max(
+            abs(added - self.last_pending_added),
+            abs(deleted - self.last_pending_deleted),
+        ) >= _LIVE_EMIT_LINE_STEP:
+            return True
+        return now - self.last_pending_at >= _LIVE_EMIT_INTERVAL_S
+
+    def mark_pending_emitted(self, added: int, deleted: int, now: float) -> None:
+        self.pending_emitted = True
+        self.last_pending_added = added
+        self.last_pending_deleted = deleted
+        self.last_pending_at = now
+
+    def matches_final_tool_call(self, tool_call: Any) -> bool:
+        call_id = getattr(tool_call, "id", None)
+        canonical = self.call_id or (self.tracker.call_id if self.tracker else "")
+        if isinstance(call_id, str) and call_id and canonical and call_id == canonical:
+            return True
+        name = getattr(tool_call, "name", None)
+        if name != self.name:
+            return False
+        arguments = getattr(tool_call, "arguments", None)
+        if not isinstance(arguments, dict):
+            return False
+        path = arguments.get("path")
+        if self.path is None and isinstance(path, str) and path:
+            self.path = path
+            return True
+        return isinstance(path, str) and path == self.path
+
+
+def _stream_key(payload: dict[str, Any]) -> str:
+    index = payload.get("index")
+    if isinstance(index, int):
+        return f"idx:{index}"
+    if isinstance(index, str) and index:
+        return f"idx:{index}"
+    call_id = payload.get("call_id")
+    if isinstance(call_id, str) and call_id:
+        return f"id:{call_id}"
+    return ""
+
+
+def _extract_complete_json_string(source: str, key: str) -> str | None:
+    match = re.search(rf'"{re.escape(key)}"\s*:\s*"', source)
+    if match is None:
+        return None
+    out: list[str] = []
+    i = match.end()
+    escape = False
+    while i < len(source):
+        ch = source[i]
+        if escape:
+            escape = False
+            if ch == "n":
+                out.append("\n")
+            elif ch == "r":
+                out.append("\r")
+            elif ch == "t":
+                out.append("\t")
+            elif ch == "u":
+                digits = source[i + 1:i + 5]
+                if len(digits) < 4:
+                    return None
+                try:
+                    out.append(chr(int(digits, 16)))
+                except ValueError:
+                    return None
+                i += 4
+            else:
+                out.append(ch)
+            i += 1
+            continue
+        if ch == "\\":
+            escape = True
+            i += 1
+            continue
+        if ch == '"':
+            return "".join(out)
+        out.append(ch)
+        i += 1
+    return None
+
+
+def _event_payload(
+    tracker: FileEditTracker,
+    *,
+    phase: str,
+    status: str,
+    added: int,
+    deleted: int,
+    approximate: bool,
+    binary: bool = False,
+) -> dict[str, Any]:
+    payload: dict[str, Any] = {
+        "version": 1,
+        "call_id": tracker.call_id,
+        "tool": tracker.tool,
+        "path": tracker.display_path,
+        "absolute_path": tracker.path.as_posix(),
+        "phase": phase,
+        "added": max(0, int(added)),
+        "deleted": max(0, int(deleted)),
+        "approximate": bool(approximate),
+        "status": status,
+    }
+    if binary:
+        payload["binary"] = True
+    return payload
+
+
+def _predict_after_text(
+    tool_name: str,
+    params: dict[str, Any],
+    before: FileSnapshot,
+) -> str | None:
+    if not before.countable:
+        return None
+    before_text = before.text or ""
+    if tool_name == "write_file":
+        content = params.get("content")
+        return content if isinstance(content, str) else ""
+    if tool_name == "edit_file":
+        old_text = params.get("old_text")
+        new_text = params.get("new_text")
+        if not isinstance(old_text, str) or not isinstance(new_text, str):
+            return None
+        replace_all = bool(params.get("replace_all"))
+        if old_text == "":
+            return new_text if not before.exists else before_text
+        if old_text in before_text:
+            if replace_all:
+                return before_text.replace(old_text, new_text)
+            return before_text.replace(old_text, new_text, 1)
+        return None
+    if tool_name == "notebook_edit":
+        return _predict_notebook_after_text(params, before_text)
+    return None
+
+
+def _predict_notebook_after_text(params: dict[str, Any], before_text: str) -> str | None:
+    try:
+        nb = json.loads(before_text) if before_text.strip() else _empty_notebook()
+    except Exception:
+        return None
+    cells = nb.get("cells")
+    if not isinstance(cells, list):
+        return None
+    try:
+        cell_index = int(params.get("cell_index", 0))
+    except (TypeError, ValueError):
+        return None
+    new_source = params.get("new_source")
+    source = new_source if isinstance(new_source, str) else ""
+    cell_type = (
+        params.get("cell_type") if params.get("cell_type") in ("code", "markdown") else "code"
+    )
+    mode = (
+        params.get("edit_mode")
+        if params.get("edit_mode") in ("replace", "insert", "delete")
+        else "replace"
+    )
+    if mode == "delete":
+        if 0 <= cell_index < len(cells):
+            cells.pop(cell_index)
+        else:
+            return None
+    elif mode == "insert":
+        insert_at = min(max(cell_index + 1, 0), len(cells))
+        cells.insert(insert_at, _new_notebook_cell(source, str(cell_type)))
+    else:
+        if not (0 <= cell_index < len(cells)):
+            return None
+        cell = cells[cell_index]
+        if not isinstance(cell, dict):
+            return None
+        cell["source"] = source
+        cell["cell_type"] = cell_type
+        if cell_type == "code":
+            cell.setdefault("outputs", [])
+            cell.setdefault("execution_count", None)
+        else:
+            cell.pop("outputs", None)
+            cell.pop("execution_count", None)
+    nb["cells"] = cells
+    try:
+        return json.dumps(nb, indent=1, ensure_ascii=False)
+    except Exception:
+        return None
+
+
+def _empty_notebook() -> dict[str, Any]:
+    return {
+        "nbformat": 4,
+        "nbformat_minor": 5,
+        "metadata": {
+            "kernelspec": {"display_name": "Python 3", "language": "python", "name": "python3"},
+            "language_info": {"name": "python"},
+        },
+        "cells": [],
+    }
+
+
+def _new_notebook_cell(source: str, cell_type: str) -> dict[str, Any]:
+    cell: dict[str, Any] = {"cell_type": cell_type, "source": source, "metadata": {}}
+    if cell_type == "code":
+        cell["outputs"] = []
+        cell["execution_count"] = None
+    return cell

nanobot/utils/helpers.py

@@ -171,6 +171,79 @@ def detect_image_mime(data: bytes) -> str | None:
     return None
 
 
+# Audio formats supported by OpenAI input_audio block
+_AUDIO_MIME_COMPAT = {"audio/wav", "audio/mpeg", "audio/mp3", "audio/aac",
+                      "audio/ogg", "audio/flac", "audio/x-m4a", "audio/mp4"}
+
+# Map MIME types to the format token expected by OpenAI-compatible input_audio APIs.
+_AUDIO_FORMAT_MAP: dict[str, str] = {
+    "audio/wav": "wav",
+    "audio/x-wav": "wav",
+    "audio/mpeg": "mp3",
+    "audio/mp3": "mp3",
+    "audio/aac": "aac",
+    "audio/ogg": "ogg",
+    "audio/flac": "flac",
+    "audio/x-m4a": "m4a",
+    "audio/mp4": "m4a",
+}
+
+
+def detect_audio_mime(data: bytes, filename: str = "") -> str | None:
+    """Detect audio MIME type from magic bytes; fallback to filename guess."""
+    if data[:4] == b"RIFF" and data[8:12] == b"WAVE":
+        return "audio/wav"
+    if data[:2] in (b"\xff\xfb", b"\xff\xf3", b"\xff\xf2", b"\xff\xfa"):
+        return "audio/mpeg"
+    if data[:4] == b"fLaC":
+        return "audio/flac"
+    if data[:4] == b"OggS":
+        return "audio/ogg"
+    if len(data) > 8 and data[4:8] == b"ftyp":
+        # Only claim audio for M4A-specific brands; avoid matching MP4 video.
+        brand = data[8:12]
+        if brand in (b"M4A ", b"M4AB", b"M4AC"):
+            return "audio/x-m4a"
+    if filename:
+        import mimetypes as _mt
+        guessed = _mt.guess_type(filename)[0]
+        if guessed and guessed.startswith("audio/"):
+            return guessed
+    return None
+
+
+def audio_mime_compat(mime: str | None) -> bool:
+    """Check if the audio MIME is compatible with OpenAI input_audio block."""
+    if not mime:
+        return False
+    return mime in _AUDIO_MIME_COMPAT
+
+
+def audio_format_for_api(mime: str) -> str:
+    """Convert an audio MIME type to the format token expected by the API.
+
+    Falls back to the subtype portion of the MIME (e.g. "x-m4a" from
+    "audio/x-m4a") when no explicit mapping exists.
+    """
+    if not mime:
+        return "wav"
+    return _AUDIO_FORMAT_MAP.get(mime, mime.split("/")[-1])
+
+
+# Video formats commonly supported by LLM APIs (data URI inline)
+_VIDEO_MIME_COMPAT = {
+    "video/mp4", "video/quicktime", "video/x-m4v",
+    "video/webm", "video/x-matroska",
+}
+
+
+def video_mime_compat(mime: str | None) -> bool:
+    """Check if the video MIME is in the commonly-supported set."""
+    if not mime:
+        return False
+    return mime in _VIDEO_MIME_COMPAT
+
+
 def build_image_content_blocks(
     raw: bytes, mime: str, path: str, label: str
 ) -> list[dict[str, Any]]:

nanobot/utils/llm_runtime.py

@@ -0,0 +1,22 @@
+"""Small helpers for passing the active LLM provider/model together."""
+
+from __future__ import annotations
+
+from collections.abc import Callable
+from dataclasses import dataclass
+
+from nanobot.providers.base import LLMProvider
+
+
+@dataclass(frozen=True)
+class LLMRuntime:
+    provider: LLMProvider
+    model: str
+
+
+LLMRuntimeResolver = Callable[[], LLMRuntime]
+
+
+def static_llm_runtime(provider: LLMProvider, model: str) -> LLMRuntimeResolver:
+    runtime = LLMRuntime(provider=provider, model=model)
+    return lambda: runtime

nanobot/utils/progress_events.py

@@ -10,13 +10,21 @@ from nanobot.agent.hook import AgentHookContext
 
 
 def on_progress_accepts_tool_events(cb: Callable[..., Any]) -> bool:
+    return _on_progress_accepts(cb, "tool_events")
+
+
+def on_progress_accepts_file_edit_events(cb: Callable[..., Any]) -> bool:
+    return _on_progress_accepts(cb, "file_edit_events")
+
+
+def _on_progress_accepts(cb: Callable[..., Any], name: str) -> bool:
     try:
         sig = inspect.signature(cb)
     except (TypeError, ValueError):
         return False
     if any(p.kind == inspect.Parameter.VAR_KEYWORD for p in sig.parameters.values()):
         return True
-    return "tool_events" in sig.parameters
+    return name in sig.parameters
 
 
 async def invoke_on_progress(
@@ -32,6 +40,15 @@ async def invoke_on_progress(
     await on_progress(content, tool_hint=tool_hint)
 
 
+async def invoke_file_edit_progress(
+    on_progress: Callable[..., Awaitable[None]],
+    file_edit_events: list[dict[str, Any]],
+) -> None:
+    if not file_edit_events or not on_progress_accepts_file_edit_events(on_progress):
+        return
+    await on_progress("", file_edit_events=file_edit_events)
+
+
 def build_tool_event_start_payload(tool_call: Any) -> dict[str, Any]:
     return {
         "version": 1,

nanobot/utils/session_attachments.py

@@ -1,74 +0,0 @@
-"""Session replay: ensure assistant ``media`` paths are under the media root.
-
-WebUI history signing (``/api/.../messages``) only works for files inside
-``get_media_dir``. Tool-driven attachments may live in the workspace; stage
-copies into the websocket media bucket before persisting message JSON.
-"""
-
-from __future__ import annotations
-
-import shutil
-import uuid
-from pathlib import Path
-from typing import Any
-
-from loguru import logger
-
-from nanobot.config.paths import get_media_dir
-from nanobot.utils.helpers import safe_filename
-
-
-def stage_media_paths_for_session_replay(paths: list[str]) -> list[str]:
-    """Keep local files only; copy anything outside the media root into ``media/websocket``."""
-    root = get_media_dir().resolve()
-    out: list[str] = []
-    seen: set[str] = set()
-    for raw in paths:
-        if not isinstance(raw, str) or not raw.strip():
-            continue
-        if raw.startswith(("http://", "https://")):
-            continue
-        try:
-            p = Path(raw).expanduser().resolve()
-        except OSError:
-            continue
-        if not p.is_file():
-            continue
-        try:
-            p.relative_to(root)
-            key = str(p)
-        except ValueError:
-            try:
-                media_dir = get_media_dir("websocket")
-                staged = media_dir / f"{uuid.uuid4().hex[:12]}-{safe_filename(p.name) or 'attachment'}"
-                shutil.copyfile(p, staged)
-                key = str(staged.resolve())
-            except OSError as exc:
-                logger.warning("failed to stage session media from {}: {}", raw, exc)
-                continue
-        if key not in seen:
-            out.append(key)
-            seen.add(key)
-    return out
-
-
-def merge_turn_media_into_last_assistant(
-    all_messages: list[dict[str, Any]],
-    generated_image_paths: list[str],
-    extra_attachment_paths: list[str],
-) -> None:
-    """Attach staged paths to the last assistant row in *all_messages* (in-place)."""
-    merged = list(
-        dict.fromkeys(
-            [
-                *stage_media_paths_for_session_replay(generated_image_paths),
-                *stage_media_paths_for_session_replay(extra_attachment_paths),
-            ]
-        )
-    )
-    last = all_messages[-1] if all_messages else None
-    if not merged or not last or last.get("role") != "assistant":
-        return
-    existing = last.get("media")
-    base = existing if isinstance(existing, list) else []
-    last["media"] = list(dict.fromkeys([*base, *merged]))

nanobot/utils/webui_titles.py

@@ -1,138 +0,0 @@
-"""Helpers for WebUI chat title generation."""
-
-from __future__ import annotations
-
-import re
-from typing import Any
-
-from loguru import logger
-
-from nanobot.providers.base import LLMProvider
-from nanobot.session.manager import Session, SessionManager
-from nanobot.utils.helpers import truncate_text
-
-WEBUI_SESSION_METADATA_KEY = "webui"
-WEBUI_TITLE_METADATA_KEY = "title"
-WEBUI_TITLE_USER_EDITED_METADATA_KEY = "title_user_edited"
-TITLE_MAX_CHARS = 60
-
-
-def mark_webui_session(session: Session, metadata: dict[str, Any]) -> bool:
-    """Persist a WebUI marker only when the inbound websocket frame opted in."""
-    if metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
-        return False
-    session.metadata[WEBUI_SESSION_METADATA_KEY] = True
-    return True
-
-
-def clean_generated_title(raw: str | None) -> str:
-    text = (raw or "").strip()
-    if not text:
-        return ""
-    text = re.sub(r"^\s*(title|标题)\s*[:：]\s*", "", text, flags=re.IGNORECASE)
-    text = text.strip().strip("\"'`“”‘’")
-    text = re.sub(r"\s+", " ", text).strip()
-    text = text.rstrip("。.!！?？,，;；:")
-    if len(text) > TITLE_MAX_CHARS:
-        text = text[: TITLE_MAX_CHARS - 1].rstrip() + "…"
-    return text
-
-
-def _title_inputs(session: Session) -> tuple[str, str]:
-    user_text = ""
-    assistant_text = ""
-    for message in session.messages:
-        role = message.get("role")
-        content = message.get("content")
-        if not isinstance(content, str) or not content.strip():
-            continue
-        if role == "user" and not user_text:
-            user_text = content.strip()
-        elif role == "assistant" and not assistant_text:
-            assistant_text = content.strip()
-        if user_text and assistant_text:
-            break
-    return user_text, assistant_text
-
-
-async def maybe_generate_webui_title(
-    *,
-    sessions: SessionManager,
-    session_key: str,
-    provider: LLMProvider,
-    model: str,
-) -> bool:
-    """Generate and persist a short title for WebUI-owned sessions only."""
-    session = sessions.get_or_create(session_key)
-    if session.metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
-        return False
-    if session.metadata.get(WEBUI_TITLE_USER_EDITED_METADATA_KEY) is True:
-        return False
-    current_title = session.metadata.get(WEBUI_TITLE_METADATA_KEY)
-    if isinstance(current_title, str) and current_title.strip():
-        return False
-
-    user_text, assistant_text = _title_inputs(session)
-    if not user_text:
-        return False
-
-    prompt = (
-        "Generate a concise title for this chat.\n"
-        "Rules:\n"
-        "- Use the same language as the user when practical.\n"
-        "- 3 to 8 words.\n"
-        "- No quotes.\n"
-        "- No punctuation at the end.\n"
-        "- Return only the title.\n\n"
-        f"User: {truncate_text(user_text, 1_000)}"
-    )
-    if assistant_text:
-        prompt += f"\nAssistant: {truncate_text(assistant_text, 1_000)}"
-
-    try:
-        response = await provider.chat_with_retry(
-            [
-                {
-                    "role": "system",
-                    "content": (
-                        "You write short, neutral chat titles. "
-                        "Return only the title text."
-                    ),
-                },
-                {"role": "user", "content": prompt},
-            ],
-            tools=None,
-            model=model,
-            max_tokens=32,
-            temperature=0.2,
-            retry_mode="standard",
-        )
-    except Exception:
-        logger.debug("Failed to generate webui session title for {}", session_key, exc_info=True)
-        return False
-
-    title = clean_generated_title(response.content)
-    if not title or title.lower().startswith("error"):
-        return False
-    session.metadata[WEBUI_TITLE_METADATA_KEY] = title
-    sessions.save(session)
-    return True
-
-
-async def maybe_generate_webui_title_after_turn(
-    *,
-    channel: str,
-    metadata: dict[str, Any],
-    sessions: SessionManager,
-    session_key: str,
-    provider: LLMProvider,
-    model: str,
-) -> bool:
-    if channel != "websocket" or metadata.get(WEBUI_SESSION_METADATA_KEY) is not True:
-        return False
-    return await maybe_generate_webui_title(
-        sessions=sessions,
-        session_key=session_key,
-        provider=provider,
-        model=model,
-    )

nanobot/utils/webui_turn_helpers.py

@@ -1,48 +0,0 @@
-"""Outbound helpers for the WebSocket/WebUI wire contract.
-
-AgentLoop uses these without importing a concrete channel plugin; only
-``channel == "websocket"`` messages are affected.
-"""
-
-from __future__ import annotations
-
-import time
-from typing import Any
-
-from nanobot.bus.events import InboundMessage, OutboundMessage
-from nanobot.bus.queue import MessageBus
-
-# Wall-clock turn start per ``chat_id`` (websocket only). Survives browser refresh while the
-# gateway process stays up; cleared on idle/stop and implicitly dropped on restart.
-_WEBSOCKET_TURN_WALL_STARTED_AT: dict[str, float] = {}
-
-
-def websocket_turn_wall_started_at(chat_id: str) -> float | None:
-    """Return ``time.time()`` when the active user turn began, if still running."""
-    return _WEBSOCKET_TURN_WALL_STARTED_AT.get(chat_id)
-
-
-async def publish_turn_run_status(bus: MessageBus, msg: InboundMessage, status: str) -> None:
-    """Notify WebSocket clients while a user turn is executing (timing strip)."""
-    if msg.channel != "websocket":
-        return
-    cid = str(msg.chat_id)
-    meta: dict[str, Any] = {
-        **dict(msg.metadata or {}),
-        "_goal_status": True,
-        "goal_status": status,
-    }
-    if status == "running":
-        t0 = time.time()
-        meta["started_at"] = t0
-        _WEBSOCKET_TURN_WALL_STARTED_AT[cid] = t0
-    else:
-        _WEBSOCKET_TURN_WALL_STARTED_AT.pop(cid, None)
-    await bus.publish_outbound(
-        OutboundMessage(
-            channel=msg.channel,
-            chat_id=cid,
-            content="",
-            metadata=meta,
-        ),
-    )

nanobot/web/__init__.py

@@ -1,6 +1,8 @@
 """Embedded web UI assets.
 
-The ``dist/`` subdirectory is populated by ``cd webui && bun run build`` and
-is shipped in the wheel; it stays empty in source checkouts until that command
-has been run.
+The ``dist/`` subdirectory holds the production WebUI bundle served by the
+gateway. It is shipped inside the published wheel and is rebuilt automatically
+by the ``webui-build`` Hatch hook during ``python -m build``. In an editable
+source checkout it stays empty until you run ``cd webui && bun run build``
+(or use the Vite dev server at ``cd webui && bun run dev``).
 """

nanobot/webui/__init__.py

@@ -0,0 +1,2 @@
+"""Backend helpers for the bundled WebUI surface."""
+

nanobot/webui/settings_api.py

@@ -0,0 +1,609 @@
+"""Settings REST helpers for the WebUI HTTP surface.
+
+The WebSocket channel owns transport/authentication. This module owns the
+settings payload shape and the allowlisted config mutations exposed to WebUI.
+"""
+
+from __future__ import annotations
+
+from typing import Any
+from zoneinfo import ZoneInfo
+
+from nanobot.config.loader import get_config_path, load_config, save_config
+from nanobot.providers.image_generation import (
+    get_image_gen_provider,
+    image_gen_provider_names,
+)
+from nanobot.providers.registry import PROVIDERS, find_by_name
+
+QueryParams = dict[str, list[str]]
+
+_WEB_SEARCH_PROVIDER_OPTIONS: tuple[dict[str, str], ...] = (
+    {"name": "duckduckgo", "label": "DuckDuckGo", "credential": "none"},
+    {"name": "brave", "label": "Brave Search", "credential": "api_key"},
+    {"name": "tavily", "label": "Tavily", "credential": "api_key"},
+    {"name": "searxng", "label": "SearXNG", "credential": "base_url"},
+    {"name": "jina", "label": "Jina", "credential": "api_key"},
+    {"name": "kagi", "label": "Kagi", "credential": "api_key"},
+    {"name": "olostep", "label": "Olostep", "credential": "api_key"},
+)
+_WEB_SEARCH_PROVIDER_BY_NAME = {
+    provider["name"]: provider for provider in _WEB_SEARCH_PROVIDER_OPTIONS
+}
+
+_IMAGE_GENERATION_ASPECT_RATIOS = {
+    "1:1",
+    "3:4",
+    "9:16",
+    "4:3",
+    "16:9",
+    "3:2",
+    "2:3",
+    "21:9",
+}
+
+
+class WebUISettingsError(ValueError):
+    """User-facing settings validation failure."""
+
+    def __init__(self, message: str, *, status: int = 400) -> None:
+        super().__init__(message)
+        self.message = message
+        self.status = status
+
+
+def _query_first(query: QueryParams, key: str) -> str | None:
+    values = query.get(key)
+    return values[0] if values else None
+
+
+def _query_first_alias(query: QueryParams, snake: str, camel: str) -> str | None:
+    value = _query_first(query, snake)
+    return _query_first(query, camel) if value is None else value
+
+
+def _mask_secret_hint(secret: str | None) -> str | None:
+    if not secret:
+        return None
+    if len(secret) <= 8:
+        return "••••"
+    return f"{secret[:4]}••••{secret[-4:]}"
+
+
+def _provider_requires_api_key(spec: Any) -> bool:
+    if spec.backend == "azure_openai":
+        return True
+    if spec.is_local or spec.is_direct:
+        return False
+    return True
+
+
+def _provider_configured_for_settings(spec: Any, provider_config: Any) -> bool:
+    if _provider_requires_api_key(spec):
+        return bool(provider_config.api_key)
+    return bool(
+        provider_config.api_key
+        or provider_config.api_base
+        or getattr(provider_config, "region", None)
+        or getattr(provider_config, "profile", None)
+    )
+
+
+def _parse_bool(value: str, field: str) -> bool:
+    normalized = value.strip().lower()
+    if normalized not in {"1", "0", "true", "false", "yes", "no"}:
+        raise WebUISettingsError(f"{field} must be boolean")
+    return normalized in {"1", "true", "yes"}
+
+
+def _image_generation_provider_rows(config: Any) -> list[dict[str, Any]]:
+    rows: list[dict[str, Any]] = []
+    for name in image_gen_provider_names():
+        spec = find_by_name(name)
+        provider_config = getattr(config.providers, name, None)
+        configured = (
+            _provider_configured_for_settings(spec, provider_config)
+            if spec is not None and provider_config is not None
+            else bool(getattr(provider_config, "api_key", None))
+        )
+        rows.append(
+            {
+                "name": name,
+                "label": spec.label if spec is not None else name,
+                "configured": configured,
+                "api_key_hint": _mask_secret_hint(
+                    getattr(provider_config, "api_key", None)
+                ),
+                "api_base": getattr(provider_config, "api_base", None),
+                "default_api_base": (
+                    spec.default_api_base if spec and spec.default_api_base else None
+                ),
+            }
+        )
+    return rows
+
+
+def settings_payload(*, requires_restart: bool = False) -> dict[str, Any]:
+    config = load_config()
+    defaults = config.agents.defaults
+    active_preset_name = defaults.model_preset or "default"
+    try:
+        effective_preset = config.resolve_preset()
+    except Exception:
+        effective_preset = config.resolve_default_preset()
+        active_preset_name = "default"
+
+    provider_name = (
+        config.get_provider_name(effective_preset.model, preset=effective_preset)
+        or effective_preset.provider
+    )
+    provider = config.get_provider(effective_preset.model, preset=effective_preset)
+    selected_provider = provider_name
+    if effective_preset.provider != "auto":
+        spec = find_by_name(effective_preset.provider)
+        selected_provider = spec.name if spec else provider_name
+
+    providers = []
+    for spec in PROVIDERS:
+        provider_config = getattr(config.providers, spec.name, None)
+        if provider_config is None or spec.is_oauth:
+            continue
+        providers.append(
+            {
+                "name": spec.name,
+                "label": spec.label,
+                "configured": _provider_configured_for_settings(spec, provider_config),
+                "api_key_required": _provider_requires_api_key(spec),
+                "api_key_hint": _mask_secret_hint(provider_config.api_key),
+                "api_base": provider_config.api_base,
+                "default_api_base": spec.default_api_base or None,
+            }
+        )
+
+    search_config = config.tools.web.search
+    image_config = config.tools.image_generation
+    search_provider = (
+        search_config.provider
+        if search_config.provider in _WEB_SEARCH_PROVIDER_BY_NAME
+        else "duckduckgo"
+    )
+    image_providers = _image_generation_provider_rows(config)
+    selected_image_provider = next(
+        (
+            provider
+            for provider in image_providers
+            if provider["name"] == image_config.provider
+        ),
+        None,
+    )
+    model_presets = [
+        {
+            "name": "default",
+            "label": "Default",
+            "active": active_preset_name == "default",
+            "is_default": True,
+            "model": defaults.model,
+            "provider": defaults.provider,
+            "max_tokens": defaults.max_tokens,
+            "context_window_tokens": defaults.context_window_tokens,
+            "temperature": defaults.temperature,
+            "reasoning_effort": defaults.reasoning_effort,
+        }
+    ]
+    for name, preset in config.model_presets.items():
+        model_presets.append(
+            {
+                "name": name,
+                "label": name,
+                "active": active_preset_name == name,
+                "is_default": False,
+                "model": preset.model,
+                "provider": preset.provider,
+                "max_tokens": preset.max_tokens,
+                "context_window_tokens": preset.context_window_tokens,
+                "temperature": preset.temperature,
+                "reasoning_effort": preset.reasoning_effort,
+            }
+        )
+
+    exec_config = config.tools.exec
+    return {
+        "agent": {
+            "model": effective_preset.model,
+            "provider": selected_provider,
+            "resolved_provider": provider_name,
+            "has_api_key": bool(provider and provider.api_key),
+            "model_preset": active_preset_name,
+            "max_tokens": effective_preset.max_tokens,
+            "context_window_tokens": effective_preset.context_window_tokens,
+            "temperature": effective_preset.temperature,
+            "reasoning_effort": effective_preset.reasoning_effort,
+            "timezone": defaults.timezone,
+            "bot_name": defaults.bot_name,
+            "bot_icon": defaults.bot_icon,
+            "tool_hint_max_length": defaults.tool_hint_max_length,
+        },
+        "model_presets": model_presets,
+        "providers": providers,
+        "web_search": {
+            "provider": search_provider,
+            "api_key_hint": _mask_secret_hint(search_config.api_key),
+            "base_url": search_config.base_url or None,
+            "max_results": search_config.max_results,
+            "timeout": search_config.timeout,
+            "providers": list(_WEB_SEARCH_PROVIDER_OPTIONS),
+        },
+        "web": {
+            "enable": config.tools.web.enable,
+            "proxy": config.tools.web.proxy,
+            "user_agent": config.tools.web.user_agent,
+            "search": {
+                "max_results": search_config.max_results,
+                "timeout": search_config.timeout,
+            },
+            "fetch": {
+                "use_jina_reader": config.tools.web.fetch.use_jina_reader,
+            },
+        },
+        "image_generation": {
+            "enabled": image_config.enabled,
+            "provider": image_config.provider,
+            "provider_configured": bool(
+                selected_image_provider and selected_image_provider["configured"]
+            ),
+            "model": image_config.model,
+            "default_aspect_ratio": image_config.default_aspect_ratio,
+            "default_image_size": image_config.default_image_size,
+            "max_images_per_turn": image_config.max_images_per_turn,
+            "save_dir": image_config.save_dir,
+            "providers": image_providers,
+        },
+        "runtime": {
+            "config_path": str(get_config_path().expanduser()),
+            "workspace_path": str(config.workspace_path),
+            "gateway_host": config.gateway.host,
+            "gateway_port": config.gateway.port,
+            "heartbeat": {
+                "enabled": config.gateway.heartbeat.enabled,
+                "interval_s": config.gateway.heartbeat.interval_s,
+                "keep_recent_messages": config.gateway.heartbeat.keep_recent_messages,
+            },
+            "dream": {
+                "schedule": defaults.dream.describe_schedule(),
+                "max_batch_size": defaults.dream.max_batch_size,
+                "max_iterations": defaults.dream.max_iterations,
+                "annotate_line_ages": defaults.dream.annotate_line_ages,
+            },
+            "unified_session": defaults.unified_session,
+        },
+        "advanced": {
+            "restrict_to_workspace": config.tools.restrict_to_workspace,
+            "ssrf_whitelist_count": len(config.tools.ssrf_whitelist),
+            "mcp_server_count": len(config.tools.mcp_servers),
+            "exec_enabled": exec_config.enable,
+            "exec_sandbox": exec_config.sandbox or None,
+            "exec_path_append_set": bool(exec_config.path_append),
+        },
+        "requires_restart": requires_restart,
+    }
+
+
+def update_agent_settings(query: QueryParams) -> dict[str, Any]:
+    config = load_config()
+    defaults = config.agents.defaults
+    changed = False
+    restart_required = False
+
+    if "model_preset" in query or "modelPreset" in query:
+        preset = (_query_first_alias(query, "model_preset", "modelPreset") or "").strip()
+        preset_value = None if not preset or preset == "default" else preset
+        if preset_value is not None and preset_value not in config.model_presets:
+            raise WebUISettingsError("unknown model preset")
+        if defaults.model_preset != preset_value:
+            defaults.model_preset = preset_value
+            changed = True
+
+    model = _query_first(query, "model")
+    if model is not None:
+        model = model.strip()
+        if not model:
+            raise WebUISettingsError("model is required")
+        if defaults.model != model:
+            defaults.model = model
+            changed = True
+
+    provider = _query_first(query, "provider")
+    if provider is not None:
+        provider = provider.strip()
+        if not provider:
+            raise WebUISettingsError("provider is required")
+        spec = find_by_name(provider)
+        if spec is None:
+            raise WebUISettingsError("unknown provider")
+        provider_config = getattr(config.providers, provider, None)
+        if (
+            provider_config is None
+            or not _provider_configured_for_settings(spec, provider_config)
+        ):
+            raise WebUISettingsError("provider is not configured")
+        if defaults.provider != provider:
+            defaults.provider = provider
+            changed = True
+
+    timezone = _query_first(query, "timezone")
+    if timezone is not None:
+        timezone = timezone.strip()
+        if not timezone:
+            raise WebUISettingsError("timezone is required")
+        try:
+            ZoneInfo(timezone)
+        except Exception:
+            raise WebUISettingsError("invalid timezone") from None
+        if defaults.timezone != timezone:
+            defaults.timezone = timezone
+            changed = True
+            restart_required = True
+
+    bot_name = _query_first_alias(query, "bot_name", "botName")
+    if bot_name is not None:
+        bot_name = bot_name.strip()
+        if not bot_name:
+            raise WebUISettingsError("bot_name is required")
+        if defaults.bot_name != bot_name:
+            defaults.bot_name = bot_name
+            changed = True
+            restart_required = True
+
+    bot_icon = _query_first_alias(query, "bot_icon", "botIcon")
+    if bot_icon is not None:
+        bot_icon = bot_icon.strip()
+        if defaults.bot_icon != bot_icon:
+            defaults.bot_icon = bot_icon
+            changed = True
+            restart_required = True
+
+    tool_hint_max_length = _query_first_alias(
+        query,
+        "tool_hint_max_length",
+        "toolHintMaxLength",
+    )
+    if tool_hint_max_length is not None:
+        try:
+            parsed = int(tool_hint_max_length)
+        except ValueError:
+            raise WebUISettingsError("tool_hint_max_length must be an integer") from None
+        if parsed < 20 or parsed > 500:
+            raise WebUISettingsError("tool_hint_max_length must be between 20 and 500")
+        if defaults.tool_hint_max_length != parsed:
+            defaults.tool_hint_max_length = parsed
+            changed = True
+            restart_required = True
+
+    if changed:
+        save_config(config)
+    return settings_payload(requires_restart=restart_required)
+
+
+def update_provider_settings(query: QueryParams) -> dict[str, Any]:
+    provider_name = (_query_first(query, "provider") or "").strip()
+    if not provider_name:
+        raise WebUISettingsError("provider is required")
+    spec = find_by_name(provider_name)
+    if spec is None or spec.is_oauth:
+        raise WebUISettingsError("unknown provider")
+
+    config = load_config()
+    provider_config = getattr(config.providers, spec.name, None)
+    if provider_config is None:
+        raise WebUISettingsError("unknown provider")
+
+    changed = False
+    if "api_key" in query or "apiKey" in query:
+        api_key = _query_first_alias(query, "api_key", "apiKey")
+        api_key = (api_key or "").strip() or None
+        if provider_config.api_key != api_key:
+            provider_config.api_key = api_key
+            changed = True
+
+    if "api_base" in query or "apiBase" in query:
+        api_base = _query_first_alias(query, "api_base", "apiBase")
+        api_base = (api_base or "").strip() or None
+        if provider_config.api_base != api_base:
+            provider_config.api_base = api_base
+            changed = True
+
+    if changed:
+        save_config(config)
+    image_config = config.tools.image_generation
+    restart_required = (
+        changed
+        and image_config.enabled
+        and image_config.provider == spec.name
+        and get_image_gen_provider(spec.name) is not None
+    )
+    return settings_payload(requires_restart=restart_required)
+
+
+def update_web_search_settings(query: QueryParams) -> dict[str, Any]:
+    provider_name = (_query_first(query, "provider") or "").strip().lower()
+    provider_option = _WEB_SEARCH_PROVIDER_BY_NAME.get(provider_name)
+    if provider_option is None:
+        raise WebUISettingsError("unknown web search provider")
+
+    config = load_config()
+    search_config = config.tools.web.search
+    web_config = config.tools.web
+    previous_provider = search_config.provider
+    changed = False
+    restart_required = False
+
+    def set_search_value(attr: str, value: object) -> None:
+        nonlocal changed
+        if getattr(search_config, attr) != value:
+            setattr(search_config, attr, value)
+            changed = True
+
+    def set_fetch_value(attr: str, value: object) -> None:
+        nonlocal changed
+        if getattr(web_config.fetch, attr) != value:
+            setattr(web_config.fetch, attr, value)
+            changed = True
+
+    if search_config.provider != provider_name:
+        search_config.provider = provider_name
+        changed = True
+
+    credential = provider_option["credential"]
+    if credential == "none":
+        set_search_value("api_key", "")
+        set_search_value("base_url", "")
+    elif credential == "base_url":
+        base_url = _query_first_alias(query, "base_url", "baseUrl")
+        base_url = base_url.strip() if base_url is not None else None
+        if not base_url and previous_provider == provider_name and search_config.base_url:
+            base_url = search_config.base_url
+        if not base_url:
+            raise WebUISettingsError("base_url is required")
+        set_search_value("base_url", base_url)
+        set_search_value("api_key", "")
+    else:
+        api_key = _query_first_alias(query, "api_key", "apiKey")
+        api_key = api_key.strip() if api_key is not None else None
+        if not api_key and previous_provider == provider_name and search_config.api_key:
+            api_key = search_config.api_key
+        if not api_key:
+            raise WebUISettingsError("api_key is required")
+        set_search_value("api_key", api_key)
+        set_search_value("base_url", "")
+
+    max_results = _query_first_alias(query, "max_results", "maxResults")
+    if max_results is not None:
+        try:
+            parsed = int(max_results)
+        except ValueError:
+            raise WebUISettingsError("max_results must be an integer") from None
+        if parsed < 1 or parsed > 10:
+            raise WebUISettingsError("max_results must be between 1 and 10")
+        set_search_value("max_results", parsed)
+
+    timeout = _query_first(query, "timeout")
+    if timeout is not None:
+        try:
+            parsed_timeout = int(timeout)
+        except ValueError:
+            raise WebUISettingsError("timeout must be an integer") from None
+        if parsed_timeout < 1 or parsed_timeout > 120:
+            raise WebUISettingsError("timeout must be between 1 and 120")
+        set_search_value("timeout", parsed_timeout)
+
+    use_jina_reader = _query_first_alias(query, "use_jina_reader", "useJinaReader")
+    if use_jina_reader is not None:
+        normalized = use_jina_reader.strip().lower()
+        if normalized not in {"1", "0", "true", "false", "yes", "no"}:
+            raise WebUISettingsError("use_jina_reader must be boolean")
+        previous_jina_reader = web_config.fetch.use_jina_reader
+        set_fetch_value("use_jina_reader", normalized in {"1", "true", "yes"})
+        if web_config.fetch.use_jina_reader != previous_jina_reader:
+            restart_required = True
+
+    if changed:
+        save_config(config)
+    return settings_payload(requires_restart=restart_required)
+
+
+def update_image_generation_settings(query: QueryParams) -> dict[str, Any]:
+    config = load_config()
+    image_config = config.tools.image_generation
+    changed = False
+
+    provider_name = _query_first(query, "provider")
+    if provider_name is not None:
+        provider_name = provider_name.strip().lower()
+        if not provider_name:
+            raise WebUISettingsError("image generation provider is required")
+        if get_image_gen_provider(provider_name) is None:
+            raise WebUISettingsError("unknown image generation provider")
+        if image_config.provider != provider_name:
+            image_config.provider = provider_name
+            changed = True
+
+    enabled = _query_first(query, "enabled")
+    if enabled is not None:
+        parsed_enabled = _parse_bool(enabled, "enabled")
+        if image_config.enabled != parsed_enabled:
+            image_config.enabled = parsed_enabled
+            changed = True
+
+    model = _query_first(query, "model")
+    if model is not None:
+        model = model.strip()
+        if not model:
+            raise WebUISettingsError("image generation model is required")
+        if len(model) > 200:
+            raise WebUISettingsError("image generation model is too long")
+        if image_config.model != model:
+            image_config.model = model
+            changed = True
+
+    default_aspect_ratio = _query_first_alias(
+        query,
+        "default_aspect_ratio",
+        "defaultAspectRatio",
+    )
+    if default_aspect_ratio is not None:
+        default_aspect_ratio = default_aspect_ratio.strip()
+        if default_aspect_ratio not in _IMAGE_GENERATION_ASPECT_RATIOS:
+            raise WebUISettingsError("unsupported image generation aspect ratio")
+        if image_config.default_aspect_ratio != default_aspect_ratio:
+            image_config.default_aspect_ratio = default_aspect_ratio
+            changed = True
+
+    default_image_size = _query_first_alias(
+        query,
+        "default_image_size",
+        "defaultImageSize",
+    )
+    if default_image_size is not None:
+        default_image_size = default_image_size.strip()
+        if not default_image_size:
+            raise WebUISettingsError("default image size is required")
+        if len(default_image_size) > 32 or not all(
+            char.isascii() and (char.isalnum() or char in {"x", "X", ":", "-", "_"})
+            for char in default_image_size
+        ):
+            raise WebUISettingsError("unsupported image generation size")
+        if image_config.default_image_size != default_image_size:
+            image_config.default_image_size = default_image_size
+            changed = True
+
+    max_images_per_turn = _query_first_alias(
+        query,
+        "max_images_per_turn",
+        "maxImagesPerTurn",
+    )
+    if max_images_per_turn is not None:
+        try:
+            parsed_max = int(max_images_per_turn)
+        except ValueError:
+            raise WebUISettingsError("max_images_per_turn must be an integer") from None
+        if parsed_max < 1 or parsed_max > 8:
+            raise WebUISettingsError("max_images_per_turn must be between 1 and 8")
+        if image_config.max_images_per_turn != parsed_max:
+            image_config.max_images_per_turn = parsed_max
+            changed = True
+
+    if image_config.enabled:
+        selected_provider = next(
+            (
+                provider
+                for provider in _image_generation_provider_rows(config)
+                if provider["name"] == image_config.provider
+            ),
+            None,
+        )
+        if not selected_provider or not selected_provider["configured"]:
+            raise WebUISettingsError("image generation provider is not configured")
+
+    if changed:
+        save_config(config)
+    return settings_payload(requires_restart=changed)

nanobot/webui/sidebar_state.py

@@ -0,0 +1,193 @@
+"""Persisted WebUI sidebar workspace state.
+
+This state is UI-only metadata, scoped to the active nanobot instance data
+directory (the directory containing the current config.json). It deliberately
+does not modify agent sessions.
+"""
+
+from __future__ import annotations
+
+import json
+import os
+import time
+from pathlib import Path
+from typing import Any
+
+from loguru import logger
+
+from nanobot.config.paths import get_webui_dir
+
+WEBUI_SIDEBAR_STATE_SCHEMA_VERSION = 1
+_MAX_STATE_FILE_BYTES = 256 * 1024
+_MAX_LIST_ITEMS = 2_000
+_MAX_MAP_ITEMS = 2_000
+_MAX_KEY_LEN = 512
+_MAX_TITLE_LEN = 160
+_MAX_TAG_LEN = 40
+_ALLOWED_DENSITIES = {"comfortable", "compact"}
+_ALLOWED_SORTS = {"updated_desc", "created_desc", "title_asc"}
+
+
+def webui_sidebar_state_path() -> Path:
+    return get_webui_dir() / "sidebar-state.json"
+
+
+def default_webui_sidebar_state() -> dict[str, Any]:
+    return {
+        "schema_version": WEBUI_SIDEBAR_STATE_SCHEMA_VERSION,
+        "pinned_keys": [],
+        "archived_keys": [],
+        "title_overrides": {},
+        "tags_by_key": {},
+        "collapsed_groups": {},
+        "view": {
+            "density": "comfortable",
+            "show_previews": False,
+            "show_timestamps": False,
+            "show_archived": False,
+            "sort": "updated_desc",
+        },
+        "updated_at": None,
+    }
+
+
+def _clean_string(value: Any, *, max_len: int = _MAX_KEY_LEN) -> str | None:
+    if not isinstance(value, str):
+        return None
+    cleaned = value.strip()
+    if not cleaned:
+        return None
+    return cleaned[:max_len]
+
+
+def _clean_string_list(value: Any, *, max_len: int = _MAX_KEY_LEN) -> list[str]:
+    if not isinstance(value, list):
+        return []
+    out: list[str] = []
+    seen: set[str] = set()
+    for item in value[:_MAX_LIST_ITEMS]:
+        cleaned = _clean_string(item, max_len=max_len)
+        if cleaned is None or cleaned in seen:
+            continue
+        seen.add(cleaned)
+        out.append(cleaned)
+    return out
+
+
+def _clean_bool_map(value: Any) -> dict[str, bool]:
+    if not isinstance(value, dict):
+        return {}
+    out: dict[str, bool] = {}
+    for key, raw in list(value.items())[:_MAX_MAP_ITEMS]:
+        cleaned_key = _clean_string(key)
+        if cleaned_key is None:
+            continue
+        out[cleaned_key] = bool(raw)
+    return out
+
+
+def _clean_title_overrides(value: Any) -> dict[str, str]:
+    if not isinstance(value, dict):
+        return {}
+    out: dict[str, str] = {}
+    for key, raw_title in list(value.items())[:_MAX_MAP_ITEMS]:
+        cleaned_key = _clean_string(key)
+        cleaned_title = _clean_string(raw_title, max_len=_MAX_TITLE_LEN)
+        if cleaned_key is None or cleaned_title is None:
+            continue
+        out[cleaned_key] = cleaned_title
+    return out
+
+
+def _clean_tags_by_key(value: Any) -> dict[str, list[str]]:
+    if not isinstance(value, dict):
+        return {}
+    out: dict[str, list[str]] = {}
+    for key, raw_tags in list(value.items())[:_MAX_MAP_ITEMS]:
+        cleaned_key = _clean_string(key)
+        if cleaned_key is None:
+            continue
+        tags = _clean_string_list(raw_tags, max_len=_MAX_TAG_LEN)[:12]
+        if tags:
+            out[cleaned_key] = tags
+    return out
+
+
+def _clean_view(value: Any) -> dict[str, Any]:
+    default = default_webui_sidebar_state()["view"]
+    if not isinstance(value, dict):
+        return dict(default)
+    density = value.get("density")
+    sort = value.get("sort")
+    return {
+        "density": density if density in _ALLOWED_DENSITIES else default["density"],
+        "show_previews": bool(value.get("show_previews", default["show_previews"])),
+        "show_timestamps": bool(value.get("show_timestamps", default["show_timestamps"])),
+        "show_archived": bool(value.get("show_archived", default["show_archived"])),
+        "sort": sort if sort in _ALLOWED_SORTS else default["sort"],
+    }
+
+
+def normalize_webui_sidebar_state(raw: Any) -> dict[str, Any]:
+    """Return a schema-v1 sidebar state from any older/partial input."""
+    if not isinstance(raw, dict):
+        raw = {}
+    state = default_webui_sidebar_state()
+    state["pinned_keys"] = _clean_string_list(raw.get("pinned_keys"))
+    state["archived_keys"] = _clean_string_list(raw.get("archived_keys"))
+    state["title_overrides"] = _clean_title_overrides(raw.get("title_overrides"))
+    state["tags_by_key"] = _clean_tags_by_key(raw.get("tags_by_key"))
+    state["collapsed_groups"] = _clean_bool_map(raw.get("collapsed_groups"))
+    state["view"] = _clean_view(raw.get("view"))
+    updated_at = raw.get("updated_at")
+    state["updated_at"] = updated_at if isinstance(updated_at, str) else None
+    return state
+
+
+def read_webui_sidebar_state() -> dict[str, Any]:
+    path = webui_sidebar_state_path()
+    if not path.is_file():
+        return default_webui_sidebar_state()
+    try:
+        if path.stat().st_size > _MAX_STATE_FILE_BYTES:
+            logger.warning("webui sidebar state too large, ignoring: {}", path)
+            return default_webui_sidebar_state()
+        with open(path, encoding="utf-8") as f:
+            raw = json.load(f)
+    except (OSError, json.JSONDecodeError) as e:
+        logger.warning("read webui sidebar state failed {}: {}", path, e)
+        return default_webui_sidebar_state()
+    return normalize_webui_sidebar_state(raw)
+
+
+def write_webui_sidebar_state(raw: dict[str, Any]) -> dict[str, Any]:
+    state = normalize_webui_sidebar_state(raw)
+    state["updated_at"] = time.strftime("%Y-%m-%dT%H:%M:%SZ", time.gmtime())
+    encoded = json.dumps(
+        state,
+        ensure_ascii=False,
+        indent=2,
+        sort_keys=True,
+    ).encode("utf-8")
+    if len(encoded) > _MAX_STATE_FILE_BYTES:
+        raise ValueError("sidebar state is too large")
+
+    path = webui_sidebar_state_path()
+    path.parent.mkdir(parents=True, exist_ok=True)
+    tmp = path.with_suffix(".json.tmp")
+    with open(tmp, "wb") as f:
+        f.write(encoded)
+        f.write(b"\n")
+        f.flush()
+        os.fsync(f.fileno())
+    os.replace(tmp, path)
+    try:
+        dir_fd = os.open(path.parent, os.O_RDONLY)
+    except OSError:
+        return state
+    try:
+        os.fsync(dir_fd)
+    finally:
+        os.close(dir_fd)
+    return state
+

nanobot/webui/thread_disk.py

@@ -1,4 +1,4 @@
-"""Legacy WebUI JSON snapshot path helpers (JSON file); transcripts use webui_transcript."""
+"""Legacy WebUI JSON snapshot path helpers (JSON file); transcripts use transcript."""
 
 from __future__ import annotations
 
@@ -8,7 +8,7 @@ from loguru import logger
 
 from nanobot.config.paths import get_webui_dir
 from nanobot.session.manager import SessionManager
-from nanobot.utils.webui_transcript import delete_webui_transcript
+from nanobot.webui.transcript import delete_webui_transcript
 
 
 def webui_thread_file_path(session_key: str) -> Path:

nanobot/webui/transcript.py

@@ -99,17 +99,39 @@ def tool_trace_lines_from_events(events: Any) -> list[str]:
     if not isinstance(events, list):
         return []
     lines: list[str] = []
+    seen: set[str] = set()
     for event in events:
         if not event or not isinstance(event, dict):
             continue
-        if event.get("phase") != "start":
+        if event.get("phase") not in {"start", "end", "error"}:
             continue
+        call_id = event.get("call_id")
+        if isinstance(call_id, str) and call_id:
+            if call_id in seen:
+                continue
+            seen.add(call_id)
         t = _format_tool_call_trace(event)
         if t:
             lines.append(t)
     return lines
 
 
+def _merge_unique_tool_trace_lines(
+    previous_traces: list[str],
+    lines: list[str],
+) -> tuple[list[str], bool]:
+    seen_lines = set(previous_traces)
+    traces = list(previous_traces)
+    added = False
+    for line in lines:
+        if line in seen_lines:
+            continue
+        seen_lines.add(line)
+        traces.append(line)
+        added = True
+    return traces, added
+
+
 def replay_transcript_to_ui_messages(
     lines: list[dict[str, Any]],
     *,
@@ -125,11 +147,36 @@ def replay_transcript_to_ui_messages(
     buffer_message_id: str | None = None
     buffer_parts: list[str] = []
     suppress_until_turn_end = False
+    active_activity_segment_id: str | None = None
+    active_file_edit_segment_id: str | None = None
+    activity_segment_counter = 0
     _ts_base = int(time.time() * 1000)
 
     def _new_id(prefix: str, idx: int) -> str:
         return f"{prefix}-{idx}-{uuid.uuid4().hex[:8]}"
 
+    def _new_activity_segment(*, activate: bool = True) -> str:
+        nonlocal active_activity_segment_id, activity_segment_counter
+        activity_segment_counter += 1
+        segment_id = f"activity-{activity_segment_counter}"
+        if activate:
+            active_activity_segment_id = segment_id
+        return segment_id
+
+    def _ensure_activity_segment() -> str:
+        return active_activity_segment_id or _new_activity_segment()
+
+    def close_activity_for_answer() -> None:
+        nonlocal active_activity_segment_id, active_file_edit_segment_id
+        active_activity_segment_id = None
+        active_file_edit_segment_id = None
+
+    def close_file_edit_phase_before_activity() -> None:
+        nonlocal active_activity_segment_id, active_file_edit_segment_id
+        if active_file_edit_segment_id:
+            active_activity_segment_id = None
+            active_file_edit_segment_id = None
+
     def attach_reasoning_chunk(prev: list[dict[str, Any]], chunk: str, idx: int) -> None:
         for i in range(len(prev) - 1, -1, -1):
             candidate = prev[i]
@@ -151,12 +198,19 @@ def replay_transcript_to_ui_messages(
                     **candidate,
                     "reasoning": (str(candidate.get("reasoning") or "")) + chunk,
                     "reasoningStreaming": True,
+                    "activitySegmentId": candidate.get("activitySegmentId") or _ensure_activity_segment(),
                 }
                 return
             if not has_answer and candidate.get("isStreaming"):
-                prev[i] = {**candidate, "reasoning": chunk, "reasoningStreaming": True}
+                prev[i] = {
+                    **candidate,
+                    "reasoning": chunk,
+                    "reasoningStreaming": True,
+                    "activitySegmentId": candidate.get("activitySegmentId") or _ensure_activity_segment(),
+                }
                 return
             break
+        segment = _ensure_activity_segment()
         prev.append(
             {
                 "id": _new_id("as", idx),
@@ -165,6 +219,7 @@ def replay_transcript_to_ui_messages(
                 "isStreaming": True,
                 "reasoning": chunk,
                 "reasoningStreaming": True,
+                "activitySegmentId": segment,
                 "createdAt": _ts_base + idx,
             },
         )
@@ -221,6 +276,7 @@ def replay_transcript_to_ui_messages(
                 return
 
     def absorb_complete(extra: dict[str, Any], idx: int) -> None:
+        nonlocal active_activity_segment_id, active_file_edit_segment_id
         last = messages[-1] if messages else None
         if last and is_reasoning_only_placeholder(last):
             messages[-1] = {
@@ -238,10 +294,98 @@ def replay_transcript_to_ui_messages(
                     **extra,
                 },
             )
+        active_activity_segment_id = None
+        active_file_edit_segment_id = None
+
+    def _file_edit_key(edit: dict[str, Any]) -> str:
+        call_id = str(edit.get("call_id") or "")
+        tool = str(edit.get("tool") or "")
+        if call_id:
+            return f"{call_id}|{tool}"
+        return f"{tool}|{edit.get('path') or ''}"
+
+    def find_file_edit_trace_index(
+        segment: str | None,
+        edits: list[dict[str, Any]],
+    ) -> int | None:
+        incoming_keys = {_file_edit_key(edit) for edit in edits if isinstance(edit, dict)}
+        for i in range(len(messages) - 1, -1, -1):
+            candidate = messages[i]
+            if candidate.get("role") == "user":
+                break
+            if candidate.get("kind") != "trace" or not candidate.get("fileEdits"):
+                continue
+            if segment and candidate.get("activitySegmentId") == segment:
+                return i
+            existing_edits = candidate.get("fileEdits")
+            if not isinstance(existing_edits, list):
+                continue
+            for existing in existing_edits:
+                if isinstance(existing, dict) and _file_edit_key(existing) in incoming_keys:
+                    return i
+        return None
+
+    def upsert_file_edits(edits: list[dict[str, Any]], idx: int) -> None:
+        nonlocal active_file_edit_segment_id
+        if not edits:
+            return
+        segment = active_file_edit_segment_id
+        target_index = find_file_edit_trace_index(segment, edits)
+        if target_index is not None:
+            last = messages[target_index]
+            segment = str(last.get("activitySegmentId") or segment or _new_activity_segment(activate=False))
+            active_file_edit_segment_id = segment
+        else:
+            if not segment:
+                segment = _new_activity_segment(activate=False)
+            active_file_edit_segment_id = segment
+            messages.append(
+                {
+                    "id": _new_id("tr", idx),
+                    "role": "tool",
+                    "kind": "trace",
+                    "content": "",
+                    "traces": [],
+                    "fileEdits": [],
+                    "activitySegmentId": segment,
+                    "createdAt": _ts_base + idx,
+                },
+            )
+            target_index = len(messages) - 1
+            last = messages[target_index]
+        if not segment:
+            segment = _new_activity_segment(activate=False)
+            active_file_edit_segment_id = segment
+        existing = list(last.get("fileEdits") or [])
+        index_by_key = {
+            _file_edit_key(edit): pos
+            for pos, edit in enumerate(existing)
+            if isinstance(edit, dict)
+        }
+        for edit in edits:
+            if not isinstance(edit, dict):
+                continue
+            key = _file_edit_key(edit)
+            if key in index_by_key:
+                pos = index_by_key[key]
+                merged = {**existing[pos], **edit}
+                if edit.get("path") and not edit.get("pending"):
+                    merged.pop("pending", None)
+                existing[pos] = merged
+            else:
+                index_by_key[key] = len(existing)
+                existing.append(dict(edit))
+        messages[target_index] = {
+            **last,
+            "fileEdits": existing,
+            "activitySegmentId": last.get("activitySegmentId") or segment,
+        }
 
     for idx, rec in enumerate(lines):
         ev = rec.get("event")
         if ev == "user":
+            active_activity_segment_id = None
+            active_file_edit_segment_id = None
             text = rec.get("text")
             text_s = text if isinstance(text, str) else ""
             media_paths = rec.get("media_paths")
@@ -264,12 +408,19 @@ def replay_transcript_to_ui_messages(
             messages.append(row)
             continue
 
+        if ev == "file_edit":
+            raw_edits = rec.get("edits")
+            if isinstance(raw_edits, list):
+                upsert_file_edits([e for e in raw_edits if isinstance(e, dict)], idx)
+            continue
+
         if ev == "delta":
             if suppress_until_turn_end:
                 continue
             chunk = rec.get("text")
             if not isinstance(chunk, str):
                 continue
+            close_activity_for_answer()
             adopted = find_active_placeholder(messages) if buffer_message_id is None else None
             if buffer_message_id is None:
                 if adopted:
@@ -308,6 +459,7 @@ def replay_transcript_to_ui_messages(
             chunk = rec.get("text")
             if not isinstance(chunk, str) or not chunk:
                 continue
+            close_file_edit_phase_before_activity()
             attach_reasoning_chunk(messages, chunk, idx)
             continue
 
@@ -329,6 +481,7 @@ def replay_transcript_to_ui_messages(
                 line = rec.get("text")
                 if not isinstance(line, str) or not line:
                     continue
+                close_file_edit_phase_before_activity()
                 attach_reasoning_chunk(messages, line, idx)
                 close_reasoning(messages)
                 continue
@@ -338,15 +491,28 @@ def replay_transcript_to_ui_messages(
                 trace_lines = structured if structured else ([text] if isinstance(text, str) and text else [])
                 if not trace_lines:
                     continue
+                segment = _ensure_activity_segment()
                 last = messages[-1] if messages else None
-                if last and last.get("kind") == "trace" and not last.get("isStreaming"):
+                if (
+                    last
+                    and last.get("kind") == "trace"
+                    and not last.get("isStreaming")
+                    and (last.get("activitySegmentId") in (None, segment))
+                ):
                     prev_traces = list(last.get("traces") or [last.get("content")])
-                    merged_traces = prev_traces + trace_lines
-                    messages[-1] = {
+                    if structured:
+                        merged_traces, added = _merge_unique_tool_trace_lines(prev_traces, structured)
+                        if not added:
+                            continue
+                    else:
+                        merged_traces = prev_traces + trace_lines
+                    merged = {
                         **last,
                         "traces": merged_traces,
-                        "content": trace_lines[-1],
+                        "content": merged_traces[-1],
+                        "activitySegmentId": last.get("activitySegmentId") or segment,
                     }
+                    messages[-1] = merged
                 else:
                     messages.append(
                         {
@@ -355,6 +521,7 @@ def replay_transcript_to_ui_messages(
                             "kind": "trace",
                             "content": trace_lines[-1],
                             "traces": trace_lines,
+                            "activitySegmentId": segment,
                             "createdAt": _ts_base + idx,
                         },
                     )
@@ -389,6 +556,8 @@ def replay_transcript_to_ui_messages(
 
         if ev == "turn_end":
             suppress_until_turn_end = False
+            active_activity_segment_id = None
+            active_file_edit_segment_id = None
             for i, m in enumerate(messages):
                 if m.get("isStreaming"):
                     messages[i] = {**m, "isStreaming": False}

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "nanobot-ai"
-version = "0.1.5.post3"
+version = "0.2.0"
 description = "A lightweight personal AI assistant framework"
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.11"
@@ -121,12 +121,22 @@ build-backend = "hatchling.build"
 [tool.hatch.metadata]
 allow-direct-references = true
 
+[tool.hatch.build.hooks.custom]
+# Implementation lives in the conventional `hatch_build.py` at the repo root.
+
 [tool.hatch.build]
 include = [
     "nanobot/**/*.py",
     "nanobot/templates/**/*.md",
     "nanobot/skills/**/*.md",
     "nanobot/skills/**/*.sh",
+    "nanobot/web/dist/**/*",
+]
+# nanobot/web/dist/ is produced by `cd webui && bun run build` and is
+# git-ignored. List it as an artifact so hatch ships it in both wheel and
+# sdist even though VCS does not track it.
+artifacts = [
+    "nanobot/web/dist/**/*",
 ]
 
 [tool.hatch.build.targets.wheel]
@@ -141,7 +151,9 @@ packages = ["nanobot"]
 [tool.hatch.build.targets.sdist]
 include = [
     "nanobot/",
+    "nanobot/web/dist/",
     "bridge/",
+    "hatch_build.py",
     "README.md",
     "LICENSE",
     "THIRD_PARTY_NOTICES.md",

tests/agent/test_auto_compact.py

@@ -45,6 +45,73 @@ def _add_turns(session, turns: int, *, prefix: str = "msg") -> None:
         session.add_message("assistant", f"{prefix} assistant {i}")
 
 
+def _make_fake_compact(
+    loop: AgentLoop,
+    *,
+    summary: str = "Summary.",
+    on_archive=None,
+    track_archived: list | None = None,
+    track_count: bool = False,
+):
+    """Return a fake compact_idle_session that mirrors the real method's session mutation."""
+    from nanobot.session.manager import Session as _Session
+
+    state = {"count": 0}
+
+    async def _fake_compact(key: str, max_suffix: int = 8) -> str:
+        state["count"] += 1
+        session = loop.sessions.get_or_create(key)
+
+        tail = list(session.messages[session.last_consolidated:])
+        if not tail:
+            session.updated_at = datetime.now()
+            loop.sessions.save(session)
+            return ""
+
+        probe = _Session(
+            key=session.key,
+            messages=tail.copy(),
+            created_at=session.created_at,
+            updated_at=session.updated_at,
+            metadata={},
+            last_consolidated=0,
+        )
+        probe.retain_recent_legal_suffix(max_suffix)
+        kept = probe.messages
+        cut = len(tail) - len(kept)
+        archive_msgs = tail[:cut]
+
+        if not archive_msgs and not kept:
+            session.updated_at = datetime.now()
+            loop.sessions.save(session)
+            return ""
+
+        last_active = session.updated_at
+        s = summary
+        if archive_msgs:
+            if on_archive:
+                result = on_archive(archive_msgs)
+                s = result if isinstance(result, str) else summary
+            if track_archived is not None:
+                track_archived.extend(archive_msgs)
+
+        if s and s != "(nothing)":
+            session.metadata["_last_summary"] = {
+                "text": s,
+                "last_active": last_active.isoformat(),
+            }
+
+        session.messages = kept
+        session.last_consolidated = 0
+        session.updated_at = datetime.now()
+        loop.sessions.save(session)
+        return s
+
+    # Attach state for count access
+    _fake_compact.state = state  # type: ignore[attr-defined]
+    return _fake_compact
+
+
 class TestSessionTTLConfig:
     """Test session TTL configuration."""
 
@@ -201,10 +268,7 @@ class TestAutoCompact:
         s2.add_message("user", "recent")
         loop.sessions.save(s2)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
         loop.auto_compact.check_expired(loop._schedule_background)
         await asyncio.sleep(0.1)
 
@@ -222,12 +286,9 @@ class TestAutoCompact:
         loop.sessions.save(session)
 
         archived_messages = []
-
-        async def _fake_archive(messages):
-            archived_messages.extend(messages)
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, track_archived=archived_messages,
+        )
 
         await loop.auto_compact._archive("cli:test")
 
@@ -246,10 +307,9 @@ class TestAutoCompact:
         _add_turns(session, 6, prefix="hello")
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "User said hello."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="User said hello.",
+        )
 
         await loop.auto_compact._archive("cli:test")
 
@@ -262,23 +322,16 @@ class TestAutoCompact:
 
     @pytest.mark.asyncio
     async def test_auto_compact_empty_session(self, tmp_path):
-        """_archive on empty session should not archive."""
+        """_archive on empty session should not store a summary."""
         loop = _make_loop(tmp_path, session_ttl_minutes=15)
 
-        archive_called = False
-
-        async def _fake_archive(messages):
-            nonlocal archive_called
-            archive_called = True
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         await loop.auto_compact._archive("cli:test")
 
-        assert not archive_called
         session_after = loop.sessions.get_or_create("cli:test")
         assert len(session_after.messages) == 0
+        assert "cli:test" not in loop.auto_compact._summaries
         await loop.close_mcp()
 
     @pytest.mark.asyncio
@@ -290,18 +343,14 @@ class TestAutoCompact:
         session.last_consolidated = 18
         loop.sessions.save(session)
 
-        archived_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archived_count
-            archived_count = len(messages)
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        archived_messages = []
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, track_archived=archived_messages,
+        )
 
         await loop.auto_compact._archive("cli:test")
 
-        assert archived_count == 2
+        assert len(archived_messages) == 2
         await loop.close_mcp()
 
 
@@ -334,12 +383,9 @@ class TestAutoCompactIdleDetection:
         loop.sessions.save(session)
 
         archived_messages = []
-
-        async def _fake_archive(messages):
-            archived_messages.extend(messages)
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, track_archived=archived_messages,
+        )
 
         # Simulate proactive archive completing before message arrives
         await loop.auto_compact._archive("cli:test")
@@ -402,10 +448,7 @@ class TestAutoCompactIdleDetection:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         msg = InboundMessage(channel="cli", sender_id="user", chat_id="test", content="/new")
         response = await loop._process_message(msg)
@@ -466,10 +509,7 @@ class TestAutoCompactSystemMessages:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         # Simulate proactive archive completing before system message arrives
         await loop.auto_compact._archive("cli:test")
@@ -547,12 +587,9 @@ class TestAutoCompactEdgeCases:
         loop.sessions.save(session)
 
         archived_messages = []
-
-        async def _fake_archive(messages):
-            archived_messages.extend(messages)
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, track_archived=archived_messages,
+        )
 
         # Simulate proactive archive completing before message arrives
         await loop.auto_compact._archive("cli:test")
@@ -644,10 +681,7 @@ class TestAutoCompactIntegration:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         # Simulate proactive archive completing before message arrives
         await loop.auto_compact._archive("cli:test")
@@ -704,12 +738,9 @@ class TestProactiveAutoCompact:
         loop.sessions.save(session)
 
         archived_messages = []
-
-        async def _fake_archive(messages):
-            archived_messages.extend(messages)
-            return "User chatted about old things."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="User chatted about old things.", track_archived=archived_messages,
+        )
 
         await self._run_check_expired(loop)
 
@@ -748,14 +779,14 @@ class TestProactiveAutoCompact:
         started = asyncio.Event()
         block_forever = asyncio.Event()
 
-        async def _slow_archive(messages):
+        async def _slow_compact(key, max_suffix=8):
             nonlocal archive_count
             archive_count += 1
             started.set()
             await block_forever.wait()
             return "Summary."
 
-        loop.consolidator.archive = _slow_archive
+        loop.consolidator.compact_idle_session = _slow_compact
 
         # First call starts archiving via callback
         loop.auto_compact.check_expired(loop._schedule_background)
@@ -781,10 +812,10 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _failing_archive(messages):
+        async def _failing_compact(key, max_suffix=8):
             raise RuntimeError("LLM down")
 
-        loop.consolidator.archive = _failing_archive
+        loop.consolidator.compact_idle_session = _failing_compact
 
         # Should not raise
         await self._run_check_expired(loop)
@@ -795,24 +826,18 @@ class TestProactiveAutoCompact:
 
     @pytest.mark.asyncio
     async def test_proactive_archive_skips_empty_sessions(self, tmp_path):
-        """Proactive archive should not call LLM for sessions with no un-consolidated messages."""
+        """Proactive archive should not produce a summary for sessions with no messages."""
         loop = _make_loop(tmp_path, session_ttl_minutes=15)
         session = loop.sessions.get_or_create("cli:test")
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_called = False
-
-        async def _fake_archive(messages):
-            nonlocal archive_called
-            archive_called = True
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         await self._run_check_expired(loop)
 
-        assert not archive_called
+        # Empty session should not produce a summary
+        assert "cli:test" not in loop.auto_compact._summaries
         await loop.close_mcp()
 
     @pytest.mark.asyncio
@@ -824,18 +849,12 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        _fake_compact = _make_fake_compact(loop)
+        loop.consolidator.compact_idle_session = _fake_compact
 
         # Simulate an active agent task for this session
         await self._run_check_expired(loop, active_session_keys={"cli:test"})
-        assert archive_count == 0
+        assert _fake_compact.state["count"] == 0
 
         session_after = loop.sessions.get_or_create("cli:test")
         assert len(session_after.messages) == 12  # All messages preserved
@@ -851,22 +870,16 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        _fake_compact = _make_fake_compact(loop)
+        loop.consolidator.compact_idle_session = _fake_compact
 
         # First tick: active task, skip
         await self._run_check_expired(loop, active_session_keys={"cli:test"})
-        assert archive_count == 0
+        assert _fake_compact.state["count"] == 0
 
         # Second tick: task completed, should archive
         await self._run_check_expired(loop)
-        assert archive_count == 1
+        assert _fake_compact.state["count"] == 1
         await loop.close_mcp()
 
     @pytest.mark.asyncio
@@ -888,18 +901,12 @@ class TestProactiveAutoCompact:
         s3.add_message("user", "recent")
         loop.sessions.save(s3)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        _fake_compact = _make_fake_compact(loop)
+        loop.consolidator.compact_idle_session = _fake_compact
 
         await self._run_check_expired(loop, active_session_keys={"cli:expired_active"})
 
-        assert archive_count == 1
+        assert _fake_compact.state["count"] == 1
         s1_after = loop.sessions.get_or_create("cli:expired_idle")
         assert len(s1_after.messages) == loop.auto_compact._RECENT_SUFFIX_MESSAGES
         s2_after = loop.sessions.get_or_create("cli:expired_active")
@@ -917,22 +924,16 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        _fake_compact = _make_fake_compact(loop)
+        loop.consolidator.compact_idle_session = _fake_compact
 
         # First tick: archives the session
         await self._run_check_expired(loop)
-        assert archive_count == 1
+        assert _fake_compact.state["count"] == 1
 
         # Second tick: should NOT re-schedule (updated_at is fresh after clear)
         await self._run_check_expired(loop)
-        assert archive_count == 1  # Still 1, not re-scheduled
+        assert _fake_compact.state["count"] == 1  # Still 1, not re-scheduled
         await loop.close_mcp()
 
     @pytest.mark.asyncio
@@ -943,22 +944,15 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         # First tick: skips (no messages), refreshes updated_at
         await self._run_check_expired(loop)
-        assert archive_count == 0
+        assert "cli:test" not in loop.auto_compact._summaries
 
         # Second tick: should NOT re-schedule because updated_at is fresh
         await self._run_check_expired(loop)
-        assert archive_count == 0
+        assert "cli:test" not in loop.auto_compact._summaries
         await loop.close_mcp()
 
     @pytest.mark.asyncio
@@ -970,18 +964,12 @@ class TestProactiveAutoCompact:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        archive_count = 0
-
-        async def _fake_archive(messages):
-            nonlocal archive_count
-            archive_count += 1
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        _fake_compact = _make_fake_compact(loop)
+        loop.consolidator.compact_idle_session = _fake_compact
 
         # First compact cycle
         await loop.auto_compact._archive("cli:test")
-        assert archive_count == 1
+        assert _fake_compact.state["count"] == 1
 
         # User returns, sends new messages
         msg = InboundMessage(channel="cli", sender_id="user", chat_id="test", content="second topic")
@@ -995,7 +983,7 @@ class TestProactiveAutoCompact:
 
         # Second compact cycle should succeed
         await loop.auto_compact._archive("cli:test")
-        assert archive_count == 2
+        assert _fake_compact.state["count"] == 2
         await loop.close_mcp()
 
 
@@ -1011,10 +999,9 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "User said hello."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="User said hello.",
+        )
 
         await loop.auto_compact._archive("cli:test")
 
@@ -1036,10 +1023,9 @@ class TestSummaryPersistence:
         session.updated_at = last_active
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "User said hello."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="User said hello.",
+        )
 
         # Archive
         await loop.auto_compact._archive("cli:test")
@@ -1069,10 +1055,7 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         await loop.auto_compact._archive("cli:test")
 
@@ -1100,10 +1083,7 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(loop)
 
         await loop.auto_compact._archive("cli:test")
 
@@ -1129,10 +1109,9 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "First summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="First summary.",
+        )
         await loop.auto_compact._archive("cli:test")
 
         # Consume the first summary via hot path
@@ -1148,10 +1127,9 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive2(messages):
-            return "Second summary."
-
-        loop.consolidator.archive = _fake_archive2
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="Second summary.",
+        )
         await loop.auto_compact._archive("cli:test")
 
         # The second archive writes a new summary
@@ -1173,10 +1151,9 @@ class TestSummaryPersistence:
         session.updated_at = datetime.now() - timedelta(minutes=20)
         loop.sessions.save(session)
 
-        async def _fake_archive(messages):
-            return "Old summary."
-
-        loop.consolidator.archive = _fake_archive
+        loop.consolidator.compact_idle_session = _make_fake_compact(
+            loop, summary="Old summary.",
+        )
         await loop.auto_compact._archive("cli:test")
 
         # Verify summary exists before /new

tests/agent/test_autocompact_unit.py

@@ -38,7 +38,7 @@ def _make_autocompact(
         sessions = MagicMock(spec=SessionManager)
     if consolidator is None:
         consolidator = MagicMock()
-        consolidator.archive = AsyncMock(return_value="Summary.")
+        consolidator.compact_idle_session = AsyncMock(return_value="Summary.")
     return AutoCompact(
         sessions=sessions,
         consolidator=consolidator,
@@ -178,62 +178,6 @@ class TestFormatSummary:
         assert result.startswith("Previous conversation summary (last active ")
 
 
-# ---------------------------------------------------------------------------
-# _split_unconsolidated
-# ---------------------------------------------------------------------------
-
-
-class TestSplitUnconsolidated:
-    """Test AutoCompact._split_unconsolidated splitting logic."""
-
-    def test_empty_session_returns_both_empty(self):
-        """Empty session should return ([], [])."""
-        ac = _make_autocompact()
-        session = _make_session(messages=[])
-        archive, kept = ac._split_unconsolidated(session)
-        assert archive == []
-        assert kept == []
-
-    def test_all_messages_archivable_when_more_than_suffix(self):
-        """Session with many messages should archive a prefix and keep suffix."""
-        ac = _make_autocompact()
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        archive, kept = ac._split_unconsolidated(session)
-        assert len(archive) > 0
-        assert len(kept) <= AutoCompact._RECENT_SUFFIX_MESSAGES
-
-    def test_fewer_messages_than_suffix_returns_empty_archive(self):
-        """Session with fewer messages than suffix should have empty archive."""
-        ac = _make_autocompact()
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(3)]
-        session = _make_session(messages=msgs)
-        archive, kept = ac._split_unconsolidated(session)
-        assert archive == []
-        assert len(kept) == len(msgs)
-
-    def test_respects_last_consolidated_offset(self):
-        """Only messages after last_consolidated should be considered."""
-        ac = _make_autocompact()
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        # First 10 are already consolidated
-        session = _make_session(messages=msgs, last_consolidated=10)
-        archive, kept = ac._split_unconsolidated(session)
-        # Only the tail of 10 messages is considered for splitting
-        assert all(m["content"] in [f"u{i}" for i in range(10, 20)] for m in kept)
-        assert all(m["content"] in [f"u{i}" for i in range(10, 20)] for m in archive)
-
-    def test_retain_recent_legal_suffix_keeps_last_n(self):
-        """The kept suffix should be at most _RECENT_SUFFIX_MESSAGES long."""
-        ac = _make_autocompact()
-        # 20 user messages = 20 messages total, all after last_consolidated=0
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        archive, kept = ac._split_unconsolidated(session)
-        assert len(kept) <= AutoCompact._RECENT_SUFFIX_MESSAGES
-        assert len(archive) == len(msgs) - len(kept)
-
-
 # ---------------------------------------------------------------------------
 # check_expired
 # ---------------------------------------------------------------------------
@@ -313,126 +257,71 @@ class TestCheckExpired:
 # ---------------------------------------------------------------------------
 
 
-class TestArchive:
-    """Test AutoCompact._archive async method."""
+class TestArchiveDelegates:
+    """_archive should delegate all session mutation to Consolidator."""
 
     @pytest.mark.asyncio
-    async def test_empty_session_updates_timestamp_no_archive_call(self):
-        """Empty session should refresh updated_at and not call consolidator.archive."""
+    async def test_calls_compact_idle_session(self):
         ac = _make_autocompact()
         mock_sm = MagicMock(spec=SessionManager)
-        empty_session = _make_session(messages=[])
-        mock_sm.get_or_create.return_value = empty_session
         ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(return_value="Summary.")
+        ac.consolidator.compact_idle_session = AsyncMock(return_value="Summary.")
 
         await ac._archive("cli:test")
 
-        ac.consolidator.archive.assert_not_called()
-        mock_sm.save.assert_called_once_with(empty_session)
-        # updated_at was refreshed
-        assert empty_session.updated_at > datetime.now() - timedelta(seconds=5)
+        ac.consolidator.compact_idle_session.assert_awaited_once_with(
+            "cli:test", ac._RECENT_SUFFIX_MESSAGES,
+        )
 
     @pytest.mark.asyncio
-    async def test_archive_returns_empty_string_no_summary_stored(self):
-        """If archive returns empty string, no summary should be stored."""
+    async def test_populates_summaries_from_metadata(self):
         ac = _make_autocompact()
         mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
+        session = _make_session(
+            metadata={"_last_summary": {"text": "Hello.", "last_active": "2026-05-13T10:00:00"}}
+        )
         mock_sm.get_or_create.return_value = session
         ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(return_value="")
+        ac.consolidator.compact_idle_session = AsyncMock(return_value="Hello.")
 
         await ac._archive("cli:test")
 
-        assert "cli:test" not in ac._summaries
-
-    @pytest.mark.asyncio
-    async def test_archive_returns_nothing_no_summary_stored(self):
-        """If archive returns '(nothing)', no summary should be stored."""
-        ac = _make_autocompact()
-        mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        mock_sm.get_or_create.return_value = session
-        ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(return_value="(nothing)")
-
-        await ac._archive("cli:test")
-
-        assert "cli:test" not in ac._summaries
-
-    @pytest.mark.asyncio
-    async def test_archive_exception_caught_key_removed_from_archiving(self):
-        """If archive raises, exception is caught and key removed from _archiving."""
-        ac = _make_autocompact()
-        mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        mock_sm.get_or_create.return_value = session
-        ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(side_effect=RuntimeError("LLM down"))
-
-        # Should not raise
-        await ac._archive("cli:test")
-
-        assert "cli:test" not in ac._archiving
-
-    @pytest.mark.asyncio
-    async def test_successful_archive_stores_summary_in_summaries_and_metadata(self):
-        """Successful archive should store summary in _summaries dict and metadata."""
-        ac = _make_autocompact()
-        mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        last_active = datetime(2026, 5, 13, 10, 0, 0)
-        session = _make_session(messages=msgs, updated_at=last_active)
-        mock_sm.get_or_create.return_value = session
-        ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(return_value="User discussed AI.")
-
-        await ac._archive("cli:test")
-
-        # _summaries
         entry = ac._summaries.get("cli:test")
         assert entry is not None
-        assert entry[0] == "User discussed AI."
-        assert entry[1] == last_active
-        # metadata
-        meta = session.metadata.get("_last_summary")
-        assert meta is not None
-        assert meta["text"] == "User discussed AI."
-        assert "last_active" in meta
+        assert entry[0] == "Hello."
 
     @pytest.mark.asyncio
-    async def test_finally_block_always_removes_from_archiving(self):
-        """Finally block should always remove key from _archiving, even on error."""
+    async def test_no_summary_when_compact_returns_empty(self):
         ac = _make_autocompact()
         mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        mock_sm.get_or_create.return_value = session
         ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(side_effect=RuntimeError("fail"))
+        ac.consolidator.compact_idle_session = AsyncMock(return_value="")
 
-        # Pre-add key to archiving to verify it gets removed
-        ac._archiving.add("cli:test")
         await ac._archive("cli:test")
-        assert "cli:test" not in ac._archiving
+
+        assert "cli:test" not in ac._summaries
 
     @pytest.mark.asyncio
-    async def test_finally_removes_from_archiving_on_success(self):
-        """Finally block should remove key from _archiving on success too."""
+    async def test_no_summary_when_compact_returns_nothing(self):
         ac = _make_autocompact()
         mock_sm = MagicMock(spec=SessionManager)
-        msgs = [{"role": "user", "content": f"u{i}"} for i in range(20)]
-        session = _make_session(messages=msgs)
-        mock_sm.get_or_create.return_value = session
         ac.sessions = mock_sm
-        ac.consolidator.archive = AsyncMock(return_value="Summary.")
+        ac.consolidator.compact_idle_session = AsyncMock(return_value="(nothing)")
+
+        await ac._archive("cli:test")
+
+        assert "cli:test" not in ac._summaries
+
+    @pytest.mark.asyncio
+    async def test_exception_still_removes_from_archiving(self):
+        ac = _make_autocompact()
+        mock_sm = MagicMock(spec=SessionManager)
+        ac.sessions = mock_sm
+        ac.consolidator.compact_idle_session = AsyncMock(side_effect=RuntimeError("fail"))
 
         ac._archiving.add("cli:test")
         await ac._archive("cli:test")
+
         assert "cli:test" not in ac._archiving
 
 

tests/agent/test_consolidator.py

@@ -28,6 +28,12 @@ def mock_provider():
 def consolidator(store, mock_provider):
     sessions = MagicMock()
     sessions.save = MagicMock()
+    # When maybe_consolidate_by_tokens refreshes the session reference via
+    # get_or_create(session.key), it should get back the same object the test
+    # passed in.  Store sessions by key so the lookup is transparent.
+    _session_cache: dict[str, MagicMock] = {}
+    sessions.get_or_create = MagicMock(side_effect=lambda key: _session_cache.get(key, MagicMock()))
+    sessions._session_cache = _session_cache
     return Consolidator(
         store=store,
         provider=mock_provider,
@@ -117,6 +123,7 @@ class TestConsolidatorTokenBudget:
         session.last_consolidated = 0
         session.messages = [{"role": "user", "content": "hi"}]
         session.key = "test:key"
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(return_value=(100, "tiktoken"))
         consolidator.archive = AsyncMock(return_value=True)
         await consolidator.maybe_consolidate_by_tokens(session)
@@ -152,6 +159,7 @@ class TestConsolidatorTokenBudget:
             session.add_message("user", f"u{i}")
             session.add_message("assistant", f"a{i}")
 
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(return_value=(100, "tiktoken"))
         consolidator.archive = AsyncMock(return_value="old conversation summary")
 
@@ -184,6 +192,7 @@ class TestConsolidatorTokenBudget:
         session.add_message("tool", "tool result", tool_call_id="call-1", name="x")
         session.add_message("assistant", "final answer")
 
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(return_value=(100, "tiktoken"))
         consolidator.archive = AsyncMock(return_value="tool turn summary")
 
@@ -210,6 +219,7 @@ class TestConsolidatorTokenBudget:
             }
             for i in range(70)
         ]
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(
             side_effect=[(1200, "tiktoken"), (400, "tiktoken")]
         )
@@ -238,6 +248,7 @@ class TestConsolidatorTokenBudget:
             for i in range(70)
         ]
         session.metadata = {}
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(
             side_effect=[(1200, "tiktoken"), (400, "tiktoken")]
         )
@@ -263,6 +274,7 @@ class TestConsolidatorTokenBudget:
             for i in range(70)
         ]
         session.metadata = {}
+        consolidator.sessions._session_cache[session.key] = session
         # Keep estimates high so the loop would otherwise run multiple rounds.
         consolidator.estimate_session_prompt_tokens = MagicMock(
             return_value=(1200, "tiktoken")
@@ -287,6 +299,7 @@ class TestConsolidatorTokenBudget:
             }
             for i in range(70)
         ]
+        consolidator.sessions._session_cache[session.key] = session
         consolidator.estimate_session_prompt_tokens = MagicMock(
             side_effect=[(1200, "tiktoken"), (400, "tiktoken")]
         )
@@ -299,6 +312,260 @@ class TestConsolidatorTokenBudget:
         assert session.last_consolidated == 61
 
 
+class TestCompactIdleSession:
+    """Tests for Consolidator.compact_idle_session — lock-protected idle truncation."""
+
+    @pytest.fixture
+    def real_consolidator(self, store, mock_provider):
+        """Create a Consolidator with a real SessionManager (not a mock)."""
+        from nanobot.session.manager import SessionManager
+
+        sessions = SessionManager(store.workspace)
+        return Consolidator(
+            store=store,
+            provider=mock_provider,
+            model="test-model",
+            sessions=sessions,
+            context_window_tokens=1000,
+            build_messages=MagicMock(return_value=[]),
+            get_tool_definitions=MagicMock(return_value=[]),
+            max_completion_tokens=100,
+        )
+
+    @pytest.mark.asyncio
+    async def test_archives_prefix_keeps_suffix(self, real_consolidator, mock_provider):
+        """20 user/assistant turns → compact with max_suffix=8 → messages ≤ 8,
+        last_consolidated=0, _last_summary stored."""
+        mock_provider.chat_with_retry.return_value = MagicMock(
+            content="Summary of old conversation.", finish_reason="stop"
+        )
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:test")
+        for i in range(20):
+            session.add_message("user", f"user msg {i}")
+            session.add_message("assistant", f"assistant msg {i}")
+        sessions.save(session)
+
+        result = await real_consolidator.compact_idle_session("cli:test", max_suffix=8)
+        assert result == "Summary of old conversation."
+
+        reloaded = sessions.get_or_create("cli:test")
+        assert len(reloaded.messages) <= 8
+        assert reloaded.last_consolidated == 0
+        meta = reloaded.metadata.get("_last_summary")
+        assert meta is not None
+        assert meta["text"] == "Summary of old conversation."
+        assert "last_active" in meta
+
+    @pytest.mark.asyncio
+    async def test_empty_session_refreshes_timestamp(self, real_consolidator):
+        """Empty session with old updated_at → refreshed after call, returns ''."""
+        from datetime import datetime, timedelta
+
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:empty")
+        old_ts = datetime.now() - timedelta(hours=2)
+        session.updated_at = old_ts
+        sessions.save(session)
+
+        result = await real_consolidator.compact_idle_session("cli:empty")
+        assert result == ""
+
+        reloaded = sessions.get_or_create("cli:empty")
+        assert reloaded.updated_at > old_ts
+
+    @pytest.mark.asyncio
+    async def test_nothing_summary_not_stored(self, real_consolidator, mock_provider):
+        """LLM returns '(nothing)' → _last_summary NOT in metadata."""
+        mock_provider.chat_with_retry.return_value = MagicMock(
+            content="(nothing)", finish_reason="stop"
+        )
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:nothing")
+        for i in range(10):
+            session.add_message("user", f"u{i}")
+            session.add_message("assistant", f"a{i}")
+        sessions.save(session)
+
+        result = await real_consolidator.compact_idle_session("cli:nothing", max_suffix=4)
+        assert result == "(nothing)"
+
+        reloaded = sessions.get_or_create("cli:nothing")
+        assert "_last_summary" not in reloaded.metadata
+
+    @pytest.mark.asyncio
+    async def test_llm_failure_still_truncates(self, real_consolidator, mock_provider, store):
+        """LLM raises RuntimeError → raw_archive fires, session still truncated, returns None."""
+        mock_provider.chat_with_retry.side_effect = RuntimeError("LLM unavailable")
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:fail")
+        for i in range(10):
+            session.add_message("user", f"u{i}")
+            session.add_message("assistant", f"a{i}")
+        sessions.save(session)
+
+        result = await real_consolidator.compact_idle_session("cli:fail", max_suffix=4)
+        assert result is None
+
+        # raw_archive should have been called (history.jsonl gets an entry)
+        entries = store.read_unprocessed_history(since_cursor=0)
+        assert any("[RAW]" in e["content"] for e in entries)
+
+        # Session should still be truncated
+        reloaded = sessions.get_or_create("cli:fail")
+        assert len(reloaded.messages) <= 4
+
+    @pytest.mark.asyncio
+    async def test_respects_last_consolidated(self, real_consolidator, mock_provider):
+        """30 turns with last_consolidated=50 → only unconsolidated tail considered."""
+        mock_provider.chat_with_retry.return_value = MagicMock(
+            content="Tail summary.", finish_reason="stop"
+        )
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:offset")
+        for i in range(30):
+            session.add_message("user", f"u{i}")
+            session.add_message("assistant", f"a{i}")
+        session.last_consolidated = 50  # Only 10 messages unconsolidated
+        sessions.save(session)
+
+        result = await real_consolidator.compact_idle_session("cli:offset", max_suffix=4)
+        assert result == "Tail summary."
+
+        # Verify only the unconsolidated tail was processed:
+        # 10 unconsolidated messages (50-59), keep suffix of 4 → archive 6
+        archived_call = mock_provider.chat_with_retry.call_args
+        user_content = archived_call.kwargs["messages"][1]["content"]
+        # Should contain only tail messages, not early ones
+        assert "u0" not in user_content
+        assert "u25" in user_content or "a25" in user_content
+
+    @pytest.mark.asyncio
+    async def test_acquires_consolidation_lock(self, real_consolidator, mock_provider):
+        """Verify lock is held during execution."""
+        import asyncio
+
+        # Use a slow LLM response to ensure the lock is held while we check
+        started = asyncio.Event()
+
+        async def slow_chat(**kwargs):
+            started.set()
+            await asyncio.sleep(0.1)
+            return MagicMock(content="Summary.", finish_reason="stop")
+
+        mock_provider.chat_with_retry = slow_chat
+
+        sessions = real_consolidator.sessions
+        session = sessions.get_or_create("cli:lock")
+        for i in range(10):
+            session.add_message("user", f"u{i}")
+            session.add_message("assistant", f"a{i}")
+        sessions.save(session)
+
+        lock = real_consolidator.get_lock("cli:lock")
+        assert not lock.locked()
+
+        task = asyncio.ensure_future(
+            real_consolidator.compact_idle_session("cli:lock", max_suffix=4)
+        )
+        await started.wait()
+        assert lock.locked()
+        await task
+        assert not lock.locked()
+
+
+class TestConsolidatorSessionRefresh:
+    """Background consolidation must detect stale session references."""
+
+    @pytest.mark.asyncio
+    async def test_reloads_before_empty_session_guard(self, tmp_path):
+        """A stale empty reference must not skip a non-empty cached session."""
+        from nanobot.agent.memory import Consolidator, MemoryStore
+        from nanobot.session.manager import Session, SessionManager
+
+        store = MemoryStore(tmp_path)
+        provider = MagicMock()
+        provider.chat_with_retry = AsyncMock(
+            return_value=MagicMock(content="summary", finish_reason="stop")
+        )
+        provider.generation.max_tokens = 4096
+        provider.estimate_prompt_tokens = MagicMock(return_value=(10, "test"))
+        sessions = SessionManager(tmp_path)
+        consolidator = Consolidator(
+            store=store,
+            provider=provider,
+            model="test-model",
+            sessions=sessions,
+            context_window_tokens=128_000,
+            build_messages=MagicMock(return_value=[]),
+            get_tool_definitions=MagicMock(return_value=[]),
+        )
+
+        fresh = sessions.get_or_create("cli:test")
+        fresh.add_message("user", "fresh message")
+        sessions.save(fresh)
+        stale_empty = Session(key="cli:test")
+
+        seen: dict[str, Session] = {}
+
+        def estimate(session: Session):
+            seen["session"] = session
+            return 10, "test"
+
+        consolidator.estimate_session_prompt_tokens = MagicMock(side_effect=estimate)
+
+        await consolidator.maybe_consolidate_by_tokens(stale_empty)
+
+        assert seen["session"] is fresh
+
+    @pytest.mark.asyncio
+    async def test_reloads_stale_session_after_compact(self, tmp_path):
+        """After compact_idle_session replaces the session, a concurrent
+        maybe_consolidate_by_tokens with the old reference should use the
+        fresh session from cache instead of overwriting."""
+        from nanobot.agent.memory import Consolidator, MemoryStore
+        from nanobot.session.manager import SessionManager
+
+        store = MemoryStore(tmp_path)
+        provider = MagicMock()
+        provider.chat_with_retry = AsyncMock(
+            return_value=MagicMock(content="summary", finish_reason="stop")
+        )
+        provider.generation.max_tokens = 4096
+        provider.estimate_prompt_tokens = MagicMock(return_value=(10, "test"))
+        sessions = SessionManager(tmp_path)
+        consolidator = Consolidator(
+            store=store,
+            provider=provider,
+            model="test-model",
+            sessions=sessions,
+            context_window_tokens=128_000,
+            build_messages=MagicMock(return_value=[]),
+            get_tool_definitions=MagicMock(return_value=[]),
+        )
+
+        # Populate session with many messages
+        session = sessions.get_or_create("cli:test")
+        for i in range(20):
+            session.add_message("user", f"u{i}")
+            session.add_message("assistant", f"a{i}")
+        sessions.save(session)
+
+        # Simulate: background consolidation captures old reference
+        old_ref = session
+
+        # AutoCompact runs first and truncates to 8
+        await consolidator.compact_idle_session("cli:test", max_suffix=8)
+
+        # Background consolidation runs with stale reference —
+        # should detect the session was replaced and not undo the compact.
+        await consolidator.maybe_consolidate_by_tokens(old_ref)
+
+        session_after = sessions.get_or_create("cli:test")
+        # Messages should still be truncated (not restored to 40)
+        assert len(session_after.messages) <= 8
+
+
 class TestRawArchiveTruncation:
     """raw_archive() must cap entry size to avoid bloating history.jsonl."""
 

tests/agent/test_context_prompt_cache.py

@@ -314,8 +314,8 @@ def test_system_prompt_keeps_message_tool_out_of_current_chat_replies(tmp_path)
     prompt = builder.build_system_prompt(channel="slack")
 
     assert "Do not use the 'message' tool for normal replies in the current chat" in prompt
-    assert "the runtime attaches those artifacts to the final assistant reply automatically" in prompt
-    assert "do not call 'message' just to announce or resend them" in prompt
+    assert "When 'generate_image' creates images" in prompt
+    assert "call 'message' with the artifact paths in the 'media' parameter" in prompt
     assert "Wait for the tool results, then answer once" in prompt
 
 

tests/agent/test_heartbeat_service.py

@@ -4,6 +4,7 @@ import pytest
 
 from nanobot.heartbeat.service import HeartbeatService
 from nanobot.providers.base import LLMProvider, LLMResponse, ToolCallRequest
+from nanobot.utils.llm_runtime import LLMRuntime
 
 
 class DummyProvider(LLMProvider):
@@ -11,9 +12,11 @@ class DummyProvider(LLMProvider):
         super().__init__()
         self._responses = list(responses)
         self.calls = 0
+        self.models: list[str | None] = []
 
     async def chat(self, *args, **kwargs) -> LLMResponse:
         self.calls += 1
+        self.models.append(kwargs.get("model"))
         if self._responses:
             return self._responses.pop(0)
         return LLMResponse(content="", tool_calls=[])
@@ -215,6 +218,51 @@ async def test_tick_suppresses_when_evaluator_says_no(tmp_path, monkeypatch) ->
     assert notified == []
 
 
+def test_tick_uses_runtime_provider_and_model(tmp_path, monkeypatch) -> None:
+    """Preset changes must apply to heartbeat decision and post-run evaluation."""
+    (tmp_path / "HEARTBEAT.md").write_text("- [ ] check runtime model", encoding="utf-8")
+
+    runtime_provider = DummyProvider([
+        LLMResponse(
+            content="",
+            tool_calls=[
+                ToolCallRequest(
+                    id="hb_1",
+                    name="heartbeat",
+                    arguments={"action": "run", "tasks": "check runtime model"},
+                )
+            ],
+        ),
+    ])
+    runtime_model = "openai/gpt-4.1"
+
+    executed: list[str] = []
+    evaluated: list[tuple[LLMProvider, str]] = []
+
+    async def _on_execute(tasks: str) -> str:
+        executed.append(tasks)
+        return "runtime model produced a user-facing update"
+
+    async def _eval_capture(response, tasks, provider, model):
+        evaluated.append((provider, model))
+        return False
+
+    service = HeartbeatService(
+        workspace=tmp_path,
+        llm_runtime=lambda: LLMRuntime(runtime_provider, runtime_model),
+        on_execute=_on_execute,
+    )
+
+    monkeypatch.setattr("nanobot.utils.evaluator.evaluate_response", _eval_capture)
+
+    asyncio.run(service._tick())
+
+    assert runtime_provider.calls == 1
+    assert runtime_provider.models == [runtime_model]
+    assert executed == ["check runtime model"]
+    assert evaluated == [(runtime_provider, runtime_model)]
+
+
 @pytest.mark.asyncio
 async def test_decide_retries_transient_error_then_succeeds(tmp_path, monkeypatch) -> None:
     provider = DummyProvider([
@@ -286,4 +334,3 @@ async def test_decide_prompt_includes_current_time(tmp_path) -> None:
     user_msg = captured_messages[1]
     assert user_msg["role"] == "user"
     assert "Current Time:" in user_msg["content"]
-

tests/agent/test_loop_image_generation_media.py

@@ -29,14 +29,15 @@ class FakeImageClient:
 
 
 @pytest.mark.asyncio
-async def test_generated_image_media_is_attached_to_final_assistant_message(
+async def test_outbound_no_longer_carries_generated_media(
     tmp_path: Path,
     monkeypatch: pytest.MonkeyPatch,
 ) -> None:
+    """Media delivery is now the LLM's responsibility via the message tool."""
     set_config_path(tmp_path / "config.json")
     monkeypatch.setattr(
-        "nanobot.agent.tools.image_generation.OpenRouterImageGenerationClient",
-        FakeImageClient,
+        "nanobot.agent.tools.image_generation.get_image_gen_provider",
+        lambda name: FakeImageClient if name == "openrouter" else None,
     )
     provider = MagicMock()
     provider.get_default_model.return_value = "test-model"
@@ -81,9 +82,6 @@ async def test_generated_image_media_is_attached_to_final_assistant_message(
 
     assert result is not None
     assert result.content == "Done"
-    assert len(result.media) == 1
-    assert Path(result.media[0]).is_file()
-
-    session = loop.sessions.get_or_create("websocket:chat-image")
-    assert session.messages[-1]["role"] == "assistant"
-    assert session.messages[-1]["media"] == result.media
+    # OutboundMessage no longer carries generated media —
+    # the LLM sends images via the message tool instead.
+    assert result.media == []

tests/agent/test_loop_progress.py

@@ -6,10 +6,15 @@ from unittest.mock import AsyncMock, MagicMock
 
 import pytest
 
+import nanobot.agent.runner as runner_module
 from nanobot.agent.loop import AgentLoop
 from nanobot.bus.events import InboundMessage
 from nanobot.bus.queue import MessageBus
 from nanobot.providers.base import LLMResponse, ToolCallRequest
+from nanobot.utils.progress_events import (
+    invoke_file_edit_progress,
+    on_progress_accepts_file_edit_events,
+)
 
 
 def _make_loop(tmp_path: Path) -> AgentLoop:
@@ -82,6 +87,143 @@ class TestToolEventProgress:
             ),
         ]
 
+    @pytest.mark.asyncio
+    async def test_write_file_emits_file_edit_progress(self, tmp_path: Path) -> None:
+        loop = _make_loop(tmp_path)
+        target = tmp_path / "foo.txt"
+        target.write_text("old\n", encoding="utf-8")
+        tool_call = ToolCallRequest(
+            id="call-write",
+            name="write_file",
+            arguments={"path": "foo.txt", "content": "new\nextra\n"},
+        )
+        calls = iter([
+            LLMResponse(content="", tool_calls=[tool_call]),
+            LLMResponse(content="Done", tool_calls=[]),
+        ])
+        loop.provider.chat_with_retry = AsyncMock(side_effect=lambda *a, **kw: next(calls))
+        loop.tools.get_definitions = MagicMock(return_value=[])
+        loop.tools.prepare_call = MagicMock(
+            return_value=(None, {"path": "foo.txt", "content": "new\nextra\n"}, None),
+        )
+
+        async def execute(name: str, params: dict) -> str:
+            target.write_text(params["content"], encoding="utf-8")
+            return "ok"
+
+        loop.tools.execute = AsyncMock(side_effect=execute)
+        file_events: list[dict] = []
+
+        async def on_progress(
+            content: str,
+            *,
+            tool_hint: bool = False,
+            tool_events: list[dict] | None = None,
+            file_edit_events: list[dict] | None = None,
+        ) -> None:
+            if file_edit_events:
+                file_events.extend(file_edit_events)
+
+        final_content, _, _, _, _ = await loop._run_agent_loop([], on_progress=on_progress)
+
+        assert final_content == "Done"
+        assert [event["phase"] for event in file_events] == ["start", "end"]
+        assert file_events[0] == {
+            "version": 1,
+            "call_id": "call-write",
+            "tool": "write_file",
+            "path": "foo.txt",
+            "absolute_path": (tmp_path / "foo.txt").resolve().as_posix(),
+            "phase": "start",
+            "added": 2,
+            "deleted": 1,
+            "approximate": True,
+            "status": "editing",
+        }
+        assert file_events[1]["status"] == "done"
+        assert file_events[1]["approximate"] is False
+        assert (file_events[1]["added"], file_events[1]["deleted"]) == (2, 1)
+
+    @pytest.mark.asyncio
+    async def test_file_edit_snapshot_skipped_when_progress_callback_cannot_emit_file_edits(
+        self,
+        tmp_path: Path,
+        monkeypatch: pytest.MonkeyPatch,
+    ) -> None:
+        loop = _make_loop(tmp_path)
+        target = tmp_path / "foo.txt"
+        target.write_text("old\n", encoding="utf-8")
+        tool_call = ToolCallRequest(
+            id="call-write",
+            name="write_file",
+            arguments={"path": "foo.txt", "content": "new\n"},
+        )
+        calls = iter([
+            LLMResponse(content="", tool_calls=[tool_call]),
+            LLMResponse(content="Done", tool_calls=[]),
+        ])
+        loop.provider.chat_with_retry = AsyncMock(side_effect=lambda *a, **kw: next(calls))
+        loop.tools.get_definitions = MagicMock(return_value=[])
+        loop.tools.prepare_call = MagicMock(
+            return_value=(None, {"path": "foo.txt", "content": "new\n"}, None),
+        )
+
+        async def execute(name: str, params: dict) -> str:
+            target.write_text(params["content"], encoding="utf-8")
+            return "ok"
+
+        loop.tools.execute = AsyncMock(side_effect=execute)
+        prepare_tracker = MagicMock(side_effect=AssertionError("unexpected file snapshot"))
+        monkeypatch.setattr(runner_module, "prepare_file_edit_tracker", prepare_tracker)
+
+        async def on_progress(
+            content: str,
+            *,
+            tool_hint: bool = False,
+            tool_events: list[dict] | None = None,
+        ) -> None:
+            pass
+
+        final_content, _, _, _, _ = await loop._run_agent_loop([], on_progress=on_progress)
+
+        assert final_content == "Done"
+        assert target.read_text(encoding="utf-8") == "new\n"
+        prepare_tracker.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_exec_does_not_emit_file_edit_progress(self, tmp_path: Path) -> None:
+        loop = _make_loop(tmp_path)
+        tool_call = ToolCallRequest(
+            id="call-exec",
+            name="exec",
+            arguments={"command": "printf hi > foo.txt"},
+        )
+        calls = iter([
+            LLMResponse(content="", tool_calls=[tool_call]),
+            LLMResponse(content="Done", tool_calls=[]),
+        ])
+        loop.provider.chat_with_retry = AsyncMock(side_effect=lambda *a, **kw: next(calls))
+        loop.tools.get_definitions = MagicMock(return_value=[])
+        loop.tools.prepare_call = MagicMock(
+            return_value=(None, {"command": "printf hi > foo.txt"}, None),
+        )
+        loop.tools.execute = AsyncMock(return_value="ok")
+        file_events: list[dict] = []
+
+        async def on_progress(
+            content: str,
+            *,
+            tool_hint: bool = False,
+            tool_events: list[dict] | None = None,
+            file_edit_events: list[dict] | None = None,
+        ) -> None:
+            if file_edit_events:
+                file_events.extend(file_edit_events)
+
+        await loop._run_agent_loop([], on_progress=on_progress)
+
+        assert file_events == []
+
     @pytest.mark.asyncio
     async def test_bus_progress_forwards_tool_events_to_outbound_metadata(self, tmp_path: Path) -> None:
         """When run() handles a bus message, _tool_events lands in OutboundMessage metadata."""
@@ -130,6 +272,138 @@ class TestToolEventProgress:
         assert finish["phase"] == "end"
         assert finish["result"] == "file.txt"
 
+    @pytest.mark.asyncio
+    async def test_bus_progress_forwards_file_edit_events_for_websocket_only(self, tmp_path: Path) -> None:
+        bus = MessageBus()
+        provider = MagicMock()
+        provider.get_default_model.return_value = "test-model"
+        loop = AgentLoop(bus=bus, provider=provider, workspace=tmp_path, model="test-model")
+        edit_events = [{
+            "call_id": "call-write",
+            "tool": "write_file",
+            "path": "foo.txt",
+            "phase": "start",
+            "added": 1,
+            "deleted": 0,
+            "approximate": True,
+            "status": "editing",
+        }]
+
+        websocket_progress = await loop._build_bus_progress_callback(InboundMessage(
+            channel="websocket",
+            sender_id="u1",
+            chat_id="chat1",
+            content="edit",
+        ))
+        assert on_progress_accepts_file_edit_events(websocket_progress) is True
+        await websocket_progress("", file_edit_events=edit_events)
+        outbound = await bus.consume_outbound()
+        assert outbound.metadata["_file_edit_events"] == edit_events
+
+        telegram_progress = await loop._build_bus_progress_callback(InboundMessage(
+            channel="telegram",
+            sender_id="u1",
+            chat_id="chat2",
+            content="edit",
+        ))
+        assert on_progress_accepts_file_edit_events(telegram_progress) is False
+        await invoke_file_edit_progress(telegram_progress, edit_events)
+        assert bus.outbound_size == 0
+
+    @pytest.mark.asyncio
+    async def test_goal_turn_keeps_live_file_edit_progress_for_webui(self, tmp_path: Path) -> None:
+        """The /goal command rewrites the prompt but must not bypass WebUI file-edit progress."""
+        bus = MessageBus()
+        provider = MagicMock()
+        provider.supports_progress_deltas = True
+        provider.get_default_model.return_value = "test-model"
+        call_count = 0
+        target = tmp_path / "goal.txt"
+
+        async def chat_stream_with_retry(*, on_tool_call_delta=None, **kwargs):
+            nonlocal call_count
+            call_count += 1
+            if call_count == 1:
+                assert on_tool_call_delta is not None
+                await on_tool_call_delta({
+                    "index": 0,
+                    "call_id": "call-goal-write",
+                    "name": "write_file",
+                    "arguments_delta": '{"path":"goal.txt","content":"',
+                })
+                await on_tool_call_delta({
+                    "index": 0,
+                    "arguments_delta": "one\\ntwo\\nthree\\n",
+                })
+                await on_tool_call_delta({"index": 0, "arguments_delta": '"}'})
+                return LLMResponse(
+                    content=None,
+                    tool_calls=[
+                        ToolCallRequest(
+                            id="call-goal-write",
+                            name="write_file",
+                            arguments={
+                                "path": "goal.txt",
+                                "content": "one\ntwo\nthree\n",
+                            },
+                        )
+                    ],
+                    usage={},
+                )
+            return LLMResponse(content="Done", tool_calls=[], usage={})
+
+        async def execute(name: str, params: dict) -> str:
+            assert name == "write_file"
+            target.write_text(params["content"], encoding="utf-8")
+            return "ok"
+
+        provider.chat_stream_with_retry = chat_stream_with_retry
+        provider.chat_with_retry = AsyncMock()
+        loop = AgentLoop(bus=bus, provider=provider, workspace=tmp_path, model="test-model")
+        loop.tools.get_definitions = MagicMock(return_value=[
+            {"type": "function", "function": {"name": "write_file"}},
+        ])
+        loop.tools.prepare_call = MagicMock(
+            return_value=(
+                None,
+                {"path": "goal.txt", "content": "one\ntwo\nthree\n"},
+                None,
+            ),
+        )
+        loop.tools.execute = AsyncMock(side_effect=execute)
+        loop.consolidator.maybe_consolidate_by_tokens = AsyncMock(return_value=False)  # type: ignore[method-assign]
+
+        await loop._dispatch(InboundMessage(
+            channel="websocket",
+            sender_id="u1",
+            chat_id="chat1",
+            content="/goal create goal file",
+            metadata={"_wants_stream": True},
+        ))
+
+        outbound = []
+        while bus.outbound_size > 0:
+            outbound.append(await bus.consume_outbound())
+
+        edit_events = [
+            event
+            for msg in outbound
+            for event in msg.metadata.get("_file_edit_events", [])
+        ]
+        assert any(
+            event["status"] == "editing"
+            and event["approximate"]
+            and event["added"] == 3
+            for event in edit_events
+        )
+        assert any(
+            event["status"] == "done"
+            and not event["approximate"]
+            and event["added"] == 3
+            for event in edit_events
+        )
+        provider.chat_with_retry.assert_not_awaited()
+
     @pytest.mark.asyncio
     async def test_non_streaming_channel_does_not_publish_codex_progress_deltas(
         self,
@@ -353,8 +627,93 @@ class TestToolEventProgress:
         assert session_updated is not None
 
         assert (session_updated.metadata or {}).get("_session_updated") is True
+        assert (session_updated.metadata or {}).get("_session_update_scope") == "metadata"
         assert provider.chat_with_retry.await_count == 2
 
+    @pytest.mark.asyncio
+    async def test_webui_title_generation_uses_turn_model_snapshot(
+        self,
+        tmp_path: Path,
+        monkeypatch: pytest.MonkeyPatch,
+    ) -> None:
+        bus = MessageBus()
+        provider = MagicMock()
+        provider.get_default_model.return_value = "test-model"
+        provider.chat_with_retry = AsyncMock(return_value=LLMResponse(content="Done", tool_calls=[]))
+        loop = AgentLoop(bus=bus, provider=provider, workspace=tmp_path, model="test-model")
+        loop.tools.get_definitions = MagicMock(return_value=[])
+        loop.consolidator.maybe_consolidate_by_tokens = AsyncMock(return_value=False)  # type: ignore[method-assign]
+
+        captured: dict[str, object] = {}
+
+        async def fake_title_after_turn(**kwargs: object) -> bool:
+            captured.update(kwargs)
+            return False
+
+        monkeypatch.setattr(
+            "nanobot.session.webui_turns.maybe_generate_webui_title_after_turn",
+            fake_title_after_turn,
+        )
+        scheduled_title: list[object] = []
+
+        def schedule_background(coro: object) -> None:
+            name = getattr(coro, "__qualname__", "")
+            if "_generate_title_and_notify" in name:
+                scheduled_title.append(coro)
+            elif hasattr(coro, "close"):
+                coro.close()
+
+        loop._schedule_background = schedule_background  # type: ignore[method-assign]
+
+        await loop._dispatch(InboundMessage(
+            channel="websocket",
+            sender_id="u1",
+            chat_id="chat1",
+            content="say hello",
+            metadata={"webui": True},
+        ))
+
+        assert len(scheduled_title) == 1
+        loop.provider = MagicMock()
+        loop.model = "switched-after-turn"
+
+        await scheduled_title[0]  # type: ignore[misc]
+
+        assert captured["provider"] is provider
+        assert captured["model"] == "test-model"
+
+    @pytest.mark.asyncio
+    async def test_webui_command_turn_does_not_schedule_title_generation(
+        self,
+        tmp_path: Path,
+        monkeypatch: pytest.MonkeyPatch,
+    ) -> None:
+        bus = MessageBus()
+        provider = MagicMock()
+        provider.get_default_model.return_value = "test-model"
+        provider.chat_with_retry = AsyncMock(return_value=LLMResponse(content="Done", tool_calls=[]))
+        loop = AgentLoop(bus=bus, provider=provider, workspace=tmp_path, model="test-model")
+
+        async def fake_title_after_turn(**_kwargs: object) -> bool:
+            raise AssertionError("command-only turns should not generate titles")
+
+        monkeypatch.setattr(
+            "nanobot.session.webui_turns.maybe_generate_webui_title_after_turn",
+            fake_title_after_turn,
+        )
+        scheduled: list[object] = []
+        loop._schedule_background = scheduled.append  # type: ignore[method-assign]
+
+        await loop._dispatch(InboundMessage(
+            channel="websocket",
+            sender_id="u1",
+            chat_id="chat1",
+            content="/model",
+            metadata={"webui": True},
+        ))
+
+        assert scheduled == []
+
     @pytest.mark.asyncio
     async def test_non_websocket_dispatch_does_not_publish_turn_end_marker(self, tmp_path: Path) -> None:
         bus = MessageBus()

tests/agent/test_loop_save_turn.py

@@ -10,12 +10,16 @@ from nanobot.bus.events import InboundMessage
 from nanobot.bus.queue import MessageBus
 from nanobot.providers.base import LLMResponse
 from nanobot.session.goal_state import GOAL_STATE_KEY
-from nanobot.session.manager import Session
-from nanobot.utils.webui_titles import (
+from nanobot.session.manager import Session, SessionManager
+from nanobot.session.webui_turns import (
+    TITLE_GENERATION_MAX_TOKENS,
+    TITLE_GENERATION_REASONING_EFFORT,
     WEBUI_SESSION_METADATA_KEY,
     WEBUI_TITLE_METADATA_KEY,
+    WebuiTurnCoordinator,
     maybe_generate_webui_title,
 )
+from nanobot.utils.llm_runtime import LLMRuntime
 
 
 def _mk_loop() -> AgentLoop:
@@ -33,6 +37,22 @@ def _make_full_loop(tmp_path: Path) -> AgentLoop:
     return AgentLoop(bus=MessageBus(), provider=provider, workspace=tmp_path, model="test-model")
 
 
+def test_agent_loop_llm_runtime_reflects_current_provider_and_model(tmp_path: Path) -> None:
+    loop = _make_full_loop(tmp_path)
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is loop.provider
+    assert runtime.model == "test-model"
+
+    next_provider = MagicMock()
+    loop.provider = next_provider
+    loop.model = "next-model"
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is next_provider
+    assert runtime.model == "next-model"
+
+
 @pytest.mark.asyncio
 async def test_generate_webui_title_only_for_marked_webui_sessions(tmp_path: Path) -> None:
     loop = _make_full_loop(tmp_path)
@@ -55,6 +75,11 @@ async def test_generate_webui_title_only_for_marked_webui_sessions(tmp_path: Pat
     assert generated is True
     assert session.metadata[WEBUI_TITLE_METADATA_KEY] == "优化 WebUI 侧边栏"
     loop.provider.chat_with_retry.assert_awaited_once()
+    assert loop.provider.chat_with_retry.await_args.kwargs["max_tokens"] == TITLE_GENERATION_MAX_TOKENS
+    assert (
+        loop.provider.chat_with_retry.await_args.kwargs["reasoning_effort"]
+        == TITLE_GENERATION_REASONING_EFFORT
+    )
 
 
 @pytest.mark.asyncio
@@ -79,6 +104,80 @@ async def test_generate_webui_title_skips_plain_websocket_sessions(tmp_path: Pat
     loop.provider.chat_with_retry.assert_not_awaited()
 
 
+@pytest.mark.asyncio
+async def test_generate_webui_title_ignores_command_only_sessions(tmp_path: Path) -> None:
+    loop = _make_full_loop(tmp_path)
+    session = loop.sessions.get_or_create("websocket:command-title")
+    session.metadata[WEBUI_SESSION_METADATA_KEY] = True
+    session.add_message("user", "/model deep", _command=True)
+    session.add_message(
+        "assistant",
+        "Switched model preset to `deep`.\n- Model: `deepseek-v4-pro`",
+        _command=True,
+    )
+    loop.sessions.save(session)
+
+    generated = await maybe_generate_webui_title(
+        sessions=loop.sessions,
+        session_key="websocket:command-title",
+        provider=loop.provider,
+        model=loop.model,
+    )
+
+    assert generated is False
+    assert WEBUI_TITLE_METADATA_KEY not in session.metadata
+    loop.provider.chat_with_retry.assert_not_awaited()
+
+
+def test_webui_title_update_uses_captured_llm_runtime(
+    tmp_path: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    bus = MessageBus()
+    sessions = SessionManager(tmp_path)
+    scheduled: list[object] = []
+    captured: dict[str, object] = {}
+
+    async def fake_title_after_turn(**kwargs: object) -> bool:
+        captured.update(kwargs)
+        return False
+
+    monkeypatch.setattr(
+        "nanobot.session.webui_turns.maybe_generate_webui_title_after_turn",
+        fake_title_after_turn,
+    )
+    coordinator = WebuiTurnCoordinator(
+        bus=bus,
+        sessions=sessions,
+        schedule_background=lambda coro: scheduled.append(coro),
+    )
+    provider = MagicMock()
+    msg = InboundMessage(
+        channel="websocket",
+        sender_id="u1",
+        chat_id="chat1",
+        content="say hello",
+        metadata={"webui": True},
+    )
+
+    coordinator.capture_title_context(
+        "websocket:chat1",
+        msg,
+        LLMRuntime(provider, "turn-model"),
+    )
+    asyncio.run(coordinator.handle_turn_end(
+        msg,
+        session_key="websocket:chat1",
+        latency_ms=None,
+    ))
+
+    assert len(scheduled) == 1
+    asyncio.run(scheduled[0])  # type: ignore[arg-type]
+
+    assert captured["provider"] is provider
+    assert captured["model"] == "turn-model"
+
+
 def test_save_turn_skips_multimodal_user_when_only_runtime_context() -> None:
     loop = _mk_loop()
     session = Session(key="test:runtime-only")

tests/agent/test_multimodal.py

@@ -0,0 +1,178 @@
+from pathlib import Path
+
+from nanobot.agent.context import ContextBuilder
+from nanobot.config.schema import InputLimitsConfig
+from nanobot.utils.helpers import detect_audio_mime, video_mime_compat
+
+PNG_BYTES = (
+    b"\x89PNG\r\n\x1a\n"
+    b"\x00\x00\x00\rIHDR"
+    b"\x00\x00\x00\x01\x00\x00\x00\x01\x08\x02\x00\x00\x00"
+    b"\x90wS\xde"
+    b"\x00\x00\x00\x0cIDATx\x9cc``\x00\x00\x00\x04\x00\x01"
+    b"\x0b\x0e-\xb4"
+    b"\x00\x00\x00\x00IEND\xaeB`\x82"
+)
+
+WAV_BYTES = b"RIFF\x24\x00\x00\x00WAVEfmt \x10\x00\x00\x00\x01\x00\x01\x00"
+
+
+def _builder(tmp_path: Path, input_limits: InputLimitsConfig | None = None) -> ContextBuilder:
+    return ContextBuilder(tmp_path, input_limits=input_limits)
+
+
+class TestAudioDetection:
+    def test_detect_wav_from_magic_bytes(self) -> None:
+        assert detect_audio_mime(WAV_BYTES) == "audio/wav"
+
+    def test_detect_mp3_from_magic_bytes(self) -> None:
+        mp3 = b"\xff\xfb\x90\x00"
+        assert detect_audio_mime(mp3) == "audio/mpeg"
+
+    def test_detect_fallback_to_filename(self) -> None:
+        assert detect_audio_mime(b"unknown", filename="song.mp3") == "audio/mpeg"
+
+    def test_returns_none_for_non_audio(self) -> None:
+        assert detect_audio_mime(PNG_BYTES) is None
+
+
+class TestVideoMimeCompat:
+    def test_mp4_is_compatible(self) -> None:
+        assert video_mime_compat("video/mp4") is True
+
+    def test_unknown_is_not_compatible(self) -> None:
+        assert video_mime_compat("video/avi") is False
+
+    def test_none_is_not_compatible(self) -> None:
+        assert video_mime_compat(None) is False
+
+
+class TestBuildUserContentMultimodal:
+    def test_audio_block_when_supported(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        path = tmp_path / "voice.wav"
+        path.write_bytes(WAV_BYTES)
+
+        content = builder._build_user_content("transcribe", [str(path)], supports_audio=True)
+
+        assert isinstance(content, list)
+        audio_blocks = [b for b in content if b.get("type") == "input_audio"]
+        assert len(audio_blocks) == 1
+        assert audio_blocks[0]["input_audio"]["format"] == "wav"
+
+    def test_audio_placeholder_when_not_supported(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        path = tmp_path / "voice.wav"
+        path.write_bytes(WAV_BYTES)
+
+        content = builder._build_user_content("transcribe", [str(path)], supports_audio=False)
+
+        assert isinstance(content, list)
+        assert any("[audio:" in b.get("text", "") for b in content)
+
+    def test_video_block_when_supported(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        path = tmp_path / "clip.mp4"
+        path.write_bytes(b"\x00" * 64)
+
+        content = builder._build_user_content("describe", [str(path)], supports_video=True)
+
+        assert isinstance(content, list)
+        video_blocks = [b for b in content if b.get("type") == "video_url"]
+        assert len(video_blocks) == 1
+        assert video_blocks[0]["video_url"]["url"].startswith("data:video/mp4;base64,")
+
+    def test_video_placeholder_when_not_supported(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        path = tmp_path / "clip.mp4"
+        path.write_bytes(b"\x00" * 64)
+
+        content = builder._build_user_content("describe", [str(path)], supports_video=False)
+
+        assert isinstance(content, list)
+        assert any("[video:" in b.get("text", "") for b in content)
+
+    def test_vision_fallback_downgrades_image(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        path = tmp_path / "pic.png"
+        path.write_bytes(PNG_BYTES)
+
+        content = builder._build_user_content("look", [str(path)], supports_vision=False)
+
+        assert isinstance(content, list)
+        assert any("[image:" in b.get("text", "") for b in content)
+        assert not any(b.get("type") == "image_url" for b in content)
+
+    def test_image_limit_count(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        paths = []
+        for i in range(5):
+            path = tmp_path / f"img{i}.png"
+            path.write_bytes(PNG_BYTES)
+            paths.append(str(path))
+
+        content = builder._build_user_content("describe", paths)
+
+        assert isinstance(content, list)
+        image_count = sum(1 for b in content if b.get("type") == "image_url")
+        assert image_count == 3  # default max_input_images
+        assert any("only the first 3 images" in b.get("text", "") for b in content)
+
+    def test_image_limit_bytes(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        big = tmp_path / "big.png"
+        big.write_bytes(PNG_BYTES + b"x" * builder.input_limits.max_input_image_bytes)
+
+        content = builder._build_user_content("analyze", [str(big)])
+
+        assert isinstance(content, str)
+        assert "file too large" in content
+
+    def test_audio_limit_count(self, tmp_path: Path) -> None:
+        limits = InputLimitsConfig(max_input_audios=1)
+        builder = _builder(tmp_path, input_limits=limits)
+        for i in range(2):
+            path = tmp_path / f"snd{i}.wav"
+            path.write_bytes(WAV_BYTES)
+
+        content = builder._build_user_content(
+            "compare", [str(tmp_path / "snd0.wav"), str(tmp_path / "snd1.wav")], supports_audio=True
+        )
+
+        assert isinstance(content, list)
+        audio_count = sum(1 for b in content if b.get("type") == "input_audio")
+        assert audio_count == 1
+        assert any("only 1 audio" in b.get("text", "") for b in content)
+
+    def test_audio_limit_bytes(self, tmp_path: Path) -> None:
+        limits = InputLimitsConfig(max_input_audio_bytes=32)
+        builder = _builder(tmp_path, input_limits=limits)
+        path = tmp_path / "big.wav"
+        path.write_bytes(WAV_BYTES + b"x" * 64)
+
+        content = builder._build_user_content("analyze", [str(path)], supports_audio=True)
+
+        assert isinstance(content, str)
+        assert "file too large" in content
+
+    def test_mixed_media_types(self, tmp_path: Path) -> None:
+        builder = _builder(tmp_path)
+        img = tmp_path / "pic.png"
+        img.write_bytes(PNG_BYTES)
+        snd = tmp_path / "voice.wav"
+        snd.write_bytes(WAV_BYTES)
+        vid = tmp_path / "clip.mp4"
+        vid.write_bytes(b"\x00" * 64)
+
+        content = builder._build_user_content(
+            "analyze",
+            [str(img), str(snd), str(vid)],
+            supports_vision=True,
+            supports_audio=True,
+            supports_video=True,
+        )
+
+        assert isinstance(content, list)
+        assert any(b.get("type") == "image_url" for b in content)
+        assert any(b.get("type") == "input_audio" for b in content)
+        assert any(b.get("type") == "video_url" for b in content)

tests/agent/test_onboard_logic.py

@@ -1074,3 +1074,242 @@ class TestConfigurePydanticModelEmptyString:
         result = _configure_pydantic_model(model, "Test")
         assert result is not None
         assert result.api_key == ""
+
+
+class TestModelPresetWizard:
+    """Tests for model preset CRUD in the onboard wizard."""
+
+    def test_sync_preset_cache(self):
+        """_sync_preset_cache should populate the module-level cache."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _sync_preset_cache
+        from nanobot.config.schema import ModelPresetConfig
+
+        config = Config()
+        config.model_presets["fast"] = ModelPresetConfig(model="gpt-4.1-mini")
+        config.model_presets["power"] = ModelPresetConfig(model="gpt-4.1")
+        _sync_preset_cache(config)
+        assert _MODEL_PRESET_CACHE == {"fast", "power"}
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_model_preset_add(self, monkeypatch):
+        """_configure_model_presets should add a new preset."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _configure_model_presets
+        from nanobot.config.schema import ModelPresetConfig
+
+        config = Config()
+        _MODEL_PRESET_CACHE.clear()
+
+        responses = iter([
+            "[+] Add new preset",
+            "my-preset",
+            "<- Back",
+        ])
+
+        class FakePrompt:
+            def __init__(self, response):
+                self.response = response
+
+            def ask(self):
+                if isinstance(self.response, BaseException):
+                    raise self.response
+                return self.response
+
+        def fake_select(*_args, **_kwargs):
+            return FakePrompt(next(responses))
+
+        def fake_text(*_args, **_kwargs):
+            return FakePrompt(next(responses))
+
+        def fake_configure(*_model, **_kwargs):
+            return ModelPresetConfig(model="gpt-test", temperature=0.5)
+
+        def fake_select_with_back(*_args, **_kwargs):
+            return next(responses)
+
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", fake_select_with_back)
+        monkeypatch.setattr(
+            onboard_wizard, "questionary", SimpleNamespace(select=fake_select, text=fake_text)
+        )
+        monkeypatch.setattr(onboard_wizard, "_configure_pydantic_model", fake_configure)
+        monkeypatch.setattr(onboard_wizard, "_show_section_header", lambda *a, **kw: None)
+        monkeypatch.setattr(onboard_wizard, "console", SimpleNamespace(clear=lambda: None))
+
+        _configure_model_presets(config)
+
+        assert "my-preset" in config.model_presets
+        assert config.model_presets["my-preset"].model == "gpt-test"
+        assert config.model_presets["my-preset"].temperature == 0.5
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_model_preset_delete(self, monkeypatch):
+        """_configure_model_presets should delete an existing preset."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _configure_model_presets
+        from nanobot.config.schema import ModelPresetConfig
+
+        config = Config()
+        config.model_presets["old"] = ModelPresetConfig(model="x")
+        _MODEL_PRESET_CACHE.clear()
+        _MODEL_PRESET_CACHE.update({"old", "default"})
+
+        responses = iter([
+            "old (x)",
+            "Delete",
+            True,
+            "<- Back",
+        ])
+
+        class FakePrompt:
+            def __init__(self, response):
+                self.response = response
+
+            def ask(self):
+                if isinstance(self.response, BaseException):
+                    raise self.response
+                return self.response
+
+        def fake_select(*_args, **_kwargs):
+            return FakePrompt(next(responses))
+
+        def fake_confirm(*_args, **_kwargs):
+            return FakePrompt(next(responses))
+
+        def fake_select_with_back(*_args, **_kwargs):
+            return next(responses)
+
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", fake_select_with_back)
+        monkeypatch.setattr(
+            onboard_wizard, "questionary", SimpleNamespace(select=fake_select, confirm=fake_confirm)
+        )
+        monkeypatch.setattr(onboard_wizard, "_show_section_header", lambda *a, **kw: None)
+        monkeypatch.setattr(onboard_wizard, "console", SimpleNamespace(clear=lambda: None))
+
+        _configure_model_presets(config)
+
+        assert "old" not in config.model_presets
+        assert "old" not in _MODEL_PRESET_CACHE
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_model_preset_field_handler(self, monkeypatch):
+        """_handle_model_preset_field should set a preset name from choices."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _handle_model_preset_field
+        from nanobot.config.schema import AgentDefaults
+
+        _MODEL_PRESET_CACHE.clear()
+        _MODEL_PRESET_CACHE.update({"fast", "power", "default"})
+
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", lambda *a, **kw: "fast")
+
+        defaults = AgentDefaults()
+        _handle_model_preset_field(defaults, "model_preset", "Model Preset", None)
+        assert defaults.model_preset == "fast"
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_model_preset_field_handler_clear(self, monkeypatch):
+        """_handle_model_preset_field should clear preset when (clear/unset) chosen."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _handle_model_preset_field
+        from nanobot.config.schema import AgentDefaults
+
+        _MODEL_PRESET_CACHE.clear()
+        _MODEL_PRESET_CACHE.add("fast")
+
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", lambda *a, **kw: "(clear/unset)")
+
+        defaults = AgentDefaults(model_preset="fast")
+        _handle_model_preset_field(defaults, "model_preset", "Model Preset", "fast")
+        assert defaults.model_preset is None
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_main_menu_dispatch_includes_model_presets(self):
+        """_configure_model_presets should be importable and callable."""
+        from nanobot.cli.onboard import _configure_model_presets
+
+        assert callable(_configure_model_presets)
+
+    def test_run_onboard_model_presets_edit(self, monkeypatch):
+        """run_onboard should handle [M] Model Presets correctly."""
+        from nanobot.config.schema import ModelPresetConfig
+
+        initial_config = Config()
+
+        responses = iter([
+            "[M] Model Presets",
+            "[S] Save and Exit",
+        ])
+
+        class FakePrompt:
+            def __init__(self, response):
+                self.response = response
+
+            def ask(self):
+                if isinstance(self.response, BaseException):
+                    raise self.response
+                return self.response
+
+        def fake_select(*_args, **_kwargs):
+            return FakePrompt(next(responses))
+
+        preset_mutated = {"n": 0}
+
+        def fake_configure_model_presets(config):
+            preset_mutated["n"] += 1
+            config.model_presets["test"] = ModelPresetConfig(model="gpt-test")
+
+        monkeypatch.setattr(onboard_wizard, "questionary", SimpleNamespace(select=fake_select))
+        monkeypatch.setattr(onboard_wizard, "_configure_model_presets", fake_configure_model_presets)
+        monkeypatch.setattr(onboard_wizard, "_show_main_menu_header", lambda: None)
+        monkeypatch.setattr(onboard_wizard, "_show_section_header", lambda *a, **kw: None)
+        monkeypatch.setattr(onboard_wizard, "console", SimpleNamespace(clear=lambda: None))
+
+        result = run_onboard(initial_config)
+        assert result.should_save is True
+        assert preset_mutated["n"] == 1
+        assert "test" in result.config.model_presets
+
+    def test_fallback_models_field_add(self, monkeypatch):
+        """_handle_fallback_models_field should add a preset name."""
+        from nanobot.cli.onboard import _MODEL_PRESET_CACHE, _handle_fallback_models_field
+        from nanobot.config.schema import AgentDefaults
+
+        _MODEL_PRESET_CACHE.clear()
+        _MODEL_PRESET_CACHE.update({"fast", "default"})
+
+        select_responses = iter(["fast"])
+        questionary_responses = iter(["[+] Add preset", "[Done]"])
+
+        class FakePrompt:
+            def __init__(self, response):
+                self.response = response
+
+            def ask(self):
+                if isinstance(self.response, BaseException):
+                    raise self.response
+                return self.response
+
+        def fake_questionary_select(*_args, **_kwargs):
+            return FakePrompt(next(questionary_responses))
+
+        def fake_select_with_back(*_args, **_kwargs):
+            return next(select_responses)
+
+        monkeypatch.setattr(
+            onboard_wizard, "questionary",
+            SimpleNamespace(select=fake_questionary_select, press_any_key_to_continue=lambda: FakePrompt(None)),
+        )
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", fake_select_with_back)
+        monkeypatch.setattr(onboard_wizard, "console", SimpleNamespace(clear=lambda: None, print=lambda *a, **kw: None))
+
+        defaults = AgentDefaults()
+        _handle_fallback_models_field(defaults, "fallback_models", "Fallback Models", [])
+        assert defaults.fallback_models == ["fast"]
+        _MODEL_PRESET_CACHE.clear()
+
+    def test_provider_field_handler(self, monkeypatch):
+        """_handle_provider_field should set provider from choices."""
+        from nanobot.cli.onboard import _handle_provider_field
+        from nanobot.config.schema import AgentDefaults
+
+        monkeypatch.setattr(onboard_wizard, "_select_with_back", lambda *a, **kw: "anthropic")
+
+        defaults = AgentDefaults()
+        _handle_provider_field(defaults, "provider", "Provider", "auto")
+        assert defaults.provider == "anthropic"

tests/agent/test_runner_progress_deltas.py

@@ -6,7 +6,7 @@ import pytest
 
 from nanobot.agent.runner import AgentRunner, AgentRunSpec
 from nanobot.config.schema import AgentDefaults
-from nanobot.providers.base import LLMResponse
+from nanobot.providers.base import LLMResponse, ToolCallRequest
 
 _MAX_TOOL_RESULT_CHARS = AgentDefaults().max_tool_result_chars
 
@@ -77,3 +77,220 @@ async def test_runner_streams_provider_progress_deltas_by_default():
     assert result.final_content == "hello"
     assert [call.args[0] for call in progress_cb.await_args_list] == ["he", "llo"]
     provider.chat_with_retry.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_runner_streams_live_write_file_activity_from_tool_argument_deltas(tmp_path):
+    provider = MagicMock()
+    provider.supports_progress_deltas = True
+    call_count = 0
+    progress_events: list[dict] = []
+
+    async def progress_cb(content, *, file_edit_events=None, **kwargs):
+        if file_edit_events:
+            progress_events.extend(file_edit_events)
+
+    class Tools:
+        def get_definitions(self):
+            return [{"type": "function", "function": {"name": "write_file"}}]
+
+        def get(self, name):
+            return None
+
+        async def execute(self, name, params):
+            assert name == "write_file"
+            assert any(event["approximate"] and event["added"] == 24 for event in progress_events)
+            target = tmp_path / params["path"]
+            target.write_text(params["content"], encoding="utf-8")
+            return "ok"
+
+    async def chat_stream_with_retry(*, on_tool_call_delta=None, **kwargs):
+        nonlocal call_count
+        call_count += 1
+        if call_count == 1:
+            assert on_tool_call_delta is not None
+            await on_tool_call_delta({
+                "index": 0,
+                "call_id": "call-write",
+                "name": "write_file",
+                "arguments_delta": '{"path":"big.txt","content":"',
+            })
+            await on_tool_call_delta({"index": 0, "arguments_delta": "line\\n" * 24})
+            return LLMResponse(
+                content=None,
+                tool_calls=[
+                    ToolCallRequest(
+                        id="call-write",
+                        name="write_file",
+                        arguments={"path": "big.txt", "content": "line\n" * 24},
+                    )
+                ],
+                usage={},
+            )
+        return LLMResponse(content="done", tool_calls=[], usage={})
+
+    provider.chat_stream_with_retry = chat_stream_with_retry
+    provider.chat_with_retry = AsyncMock()
+
+    runner = AgentRunner(provider)
+    result = await runner.run(AgentRunSpec(
+        initial_messages=[{"role": "user", "content": "write a large file"}],
+        tools=Tools(),
+        model="test-model",
+        max_iterations=2,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+        progress_callback=progress_cb,
+        workspace=tmp_path,
+    ))
+
+    assert result.final_content == "done"
+    assert any(event["approximate"] and event["added"] == 24 for event in progress_events)
+    assert any(
+        not event["approximate"] and event["phase"] == "end" and event["added"] == 24
+        for event in progress_events
+    )
+    provider.chat_with_retry.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_runner_streams_live_edit_file_activity_from_tool_argument_deltas(tmp_path):
+    provider = MagicMock()
+    provider.supports_progress_deltas = True
+    call_count = 0
+    progress_events: list[dict] = []
+    target = tmp_path / "notes.txt"
+    target.write_text("old\nkeep\n", encoding="utf-8")
+
+    async def progress_cb(content, *, file_edit_events=None, **kwargs):
+        if file_edit_events:
+            progress_events.extend(file_edit_events)
+
+    class Tools:
+        def get_definitions(self):
+            return [{"type": "function", "function": {"name": "edit_file"}}]
+
+        def get(self, name):
+            return None
+
+        async def execute(self, name, params):
+            assert name == "edit_file"
+            assert any(
+                event["tool"] == "edit_file"
+                and event["approximate"]
+                and event["added"] == 3
+                and event["deleted"] == 2
+                for event in progress_events
+            )
+            target.write_text(params["new_text"], encoding="utf-8")
+            return "ok"
+
+    async def chat_stream_with_retry(*, on_tool_call_delta=None, **kwargs):
+        nonlocal call_count
+        call_count += 1
+        if call_count == 1:
+            assert on_tool_call_delta is not None
+            await on_tool_call_delta({
+                "index": 0,
+                "call_id": "call-edit",
+                "name": "edit_file",
+                "arguments_delta": (
+                    '{"path":"notes.txt","old_text":"old\\nkeep\\n","new_text":"'
+                ),
+            })
+            await on_tool_call_delta({
+                "index": 0,
+                "arguments_delta": "new\\nkeep\\nextra\\n",
+            })
+            await on_tool_call_delta({"index": 0, "arguments_delta": '"}'})
+            return LLMResponse(
+                content=None,
+                tool_calls=[
+                    ToolCallRequest(
+                        id="call-edit",
+                        name="edit_file",
+                        arguments={
+                            "path": "notes.txt",
+                            "old_text": "old\nkeep\n",
+                            "new_text": "new\nkeep\nextra\n",
+                        },
+                    )
+                ],
+                usage={},
+            )
+        return LLMResponse(content="done", tool_calls=[], usage={})
+
+    provider.chat_stream_with_retry = chat_stream_with_retry
+    provider.chat_with_retry = AsyncMock()
+
+    runner = AgentRunner(provider)
+    result = await runner.run(AgentRunSpec(
+        initial_messages=[{"role": "user", "content": "edit a file"}],
+        tools=Tools(),
+        model="test-model",
+        max_iterations=2,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+        progress_callback=progress_cb,
+        workspace=tmp_path,
+    ))
+
+    assert result.final_content == "done"
+    assert any(
+        event["tool"] == "edit_file"
+        and event["approximate"]
+        and event["added"] == 3
+        and event["deleted"] == 2
+        for event in progress_events
+    )
+    assert any(
+        event["tool"] == "edit_file"
+        and not event["approximate"]
+        and event["phase"] == "end"
+        and event["added"] == 2
+        and event["deleted"] == 1
+        for event in progress_events
+    )
+    provider.chat_with_retry.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_runner_marks_unfinished_live_write_file_activity_failed(tmp_path):
+    provider = MagicMock()
+    provider.supports_progress_deltas = True
+    progress_events: list[dict] = []
+
+    async def progress_cb(content, *, file_edit_events=None, **kwargs):
+        if file_edit_events:
+            progress_events.extend(file_edit_events)
+
+    async def chat_stream_with_retry(*, on_tool_call_delta=None, **kwargs):
+        assert on_tool_call_delta is not None
+        await on_tool_call_delta({
+            "index": 0,
+            "call_id": "call-write",
+            "name": "write_file",
+            "arguments_delta": '{"path":"aborted.txt","content":"partial\\n',
+        })
+        return LLMResponse(content="stopped", tool_calls=[], finish_reason="stop", usage={})
+
+    provider.chat_stream_with_retry = chat_stream_with_retry
+    provider.chat_with_retry = AsyncMock()
+    tools = MagicMock()
+    tools.get_definitions.return_value = [{"type": "function", "function": {"name": "write_file"}}]
+    tools.get.return_value = None
+
+    runner = AgentRunner(provider)
+    result = await runner.run(AgentRunSpec(
+        initial_messages=[{"role": "user", "content": "write a large file"}],
+        tools=tools,
+        model="test-model",
+        max_iterations=1,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+        progress_callback=progress_cb,
+        workspace=tmp_path,
+    ))
+
+    assert result.final_content == "stopped"
+    assert progress_events[-1]["path"] == "aborted.txt"
+    assert progress_events[-1]["phase"] == "error"
+    assert progress_events[-1]["status"] == "error"
+    provider.chat_with_retry.assert_not_awaited()

tests/agent/test_runtime_refresh.py

@@ -47,3 +47,28 @@ def test_provider_refresh_updates_all_model_dependents(tmp_path: Path) -> None:
     assert loop.dream.provider is new_provider
     assert loop.dream.model == "new-model"
     assert loop.dream._runner.provider is new_provider
+
+
+def test_llm_runtime_refreshes_provider_snapshot(tmp_path: Path) -> None:
+    old_provider = _provider("old-model")
+    new_provider = _provider("new-model", max_tokens=456)
+    loop = AgentLoop(
+        bus=MessageBus(),
+        provider=old_provider,
+        workspace=tmp_path,
+        model="old-model",
+        context_window_tokens=1000,
+        provider_snapshot_loader=lambda: ProviderSnapshot(
+            provider=new_provider,
+            model="new-model",
+            context_window_tokens=2000,
+            signature=("new-model",),
+        ),
+    )
+
+    runtime = loop.llm_runtime()
+
+    assert runtime.provider is new_provider
+    assert runtime.model == "new-model"
+    assert loop.provider is new_provider
+    assert loop.runner.provider is new_provider

tests/agent/test_session_media_persist.py

@@ -1,34 +0,0 @@
-"""Tests for staging attachment paths into the media bucket for session replay."""
-
-from pathlib import Path
-
-from nanobot.config.loader import set_config_path
-from nanobot.config.paths import get_media_dir
-from nanobot.utils.session_attachments import stage_media_paths_for_session_replay
-
-
-def test_persist_media_stages_workspace_file(tmp_path: Path) -> None:
-    set_config_path(tmp_path / "config.json")
-    outside = tmp_path / "workspace" / "report.md"
-    outside.parent.mkdir(parents=True)
-    outside.write_text("body", encoding="utf-8")
-
-    out = stage_media_paths_for_session_replay([str(outside)])
-
-    assert len(out) == 1
-    staged = Path(out[0])
-    assert staged.is_file()
-    assert staged.read_text(encoding="utf-8") == "body"
-    assert staged.resolve().is_relative_to(get_media_dir().resolve())
-
-
-def test_persist_media_keeps_files_already_under_media_root(tmp_path: Path) -> None:
-    set_config_path(tmp_path / "config.json")
-    media = get_media_dir("websocket")
-    media.mkdir(parents=True, exist_ok=True)
-    inside = media / "keep-me.txt"
-    inside.write_text("x", encoding="utf-8")
-
-    out = stage_media_paths_for_session_replay([str(inside.resolve())])
-
-    assert out == [str(inside.resolve())]

tests/agent/test_unified_session.py

@@ -387,6 +387,7 @@ class TestConsolidationUnaffectedByUnifiedSession:
 
         session = Session(key="unified:default")
         session.messages = [{"role": "user", "content": "msg"}]
+        sessions.get_or_create.return_value = session
 
         # Simulate over-budget: estimated > budget
         consolidator.estimate_session_prompt_tokens = MagicMock(return_value=(950, "tiktoken"))

tests/channels/test_feishu_post_content.py

@@ -27,10 +27,11 @@ def test_extract_post_content_supports_post_wrapper_shape() -> None:
         }
     }
 
-    text, image_keys = _extract_post_content(payload)
+    text, image_keys, media_items = _extract_post_content(payload)
 
     assert text == "日报 完成"
     assert image_keys == ["img_1"]
+    assert media_items == []
 
 
 def test_extract_post_content_keeps_direct_shape_behavior() -> None:
@@ -45,10 +46,29 @@ def test_extract_post_content_keeps_direct_shape_behavior() -> None:
         ],
     }
 
-    text, image_keys = _extract_post_content(payload)
+    text, image_keys, media_items = _extract_post_content(payload)
 
     assert text == "Daily report"
     assert image_keys == ["img_a", "img_b"]
+    assert media_items == []
+
+
+def test_extract_post_content_extracts_media_tags() -> None:
+    payload = {
+        "title": "Video",
+        "content": [
+            [
+                {"tag": "text", "text": "see this"},
+                {"tag": "media", "file_key": "vid_1"},
+            ]
+        ],
+    }
+
+    text, image_keys, media_items = _extract_post_content(payload)
+
+    assert text == "Video see this"
+    assert image_keys == []
+    assert media_items == [{"tag": "media", "file_key": "vid_1"}]
 
 
 def test_register_optional_event_keeps_builder_when_method_missing() -> None:

tests/channels/test_websocket_channel.py

@@ -29,7 +29,8 @@ from nanobot.channels.websocket import (
     publish_runtime_model_update,
 )
 from nanobot.config.loader import load_config, save_config
-from nanobot.config.schema import Config
+from nanobot.config.schema import Config, ModelPresetConfig
+from nanobot.webui.settings_api import settings_payload
 
 # -- Shared helpers (aligned with test_websocket_integration.py) ---------------
 
@@ -370,6 +371,55 @@ async def test_send_progress_includes_structured_tool_events() -> None:
     ]
 
 
+@pytest.mark.asyncio
+async def test_send_file_edit_progress_uses_file_edit_event() -> None:
+    bus = MagicMock()
+    channel = WebSocketChannel({"enabled": True, "allowFrom": ["*"]}, bus)
+    mock_ws = AsyncMock()
+    channel._attach(mock_ws, "chat-1")
+
+    await channel.send(OutboundMessage(
+        channel="websocket",
+        chat_id="chat-1",
+        content="",
+        metadata={
+            "_progress": True,
+            "_file_edit_events": [
+                {
+                    "version": 1,
+                    "phase": "start",
+                    "call_id": "call-1",
+                    "tool": "write_file",
+                    "path": "src/app.py",
+                    "added": 12,
+                    "deleted": 2,
+                    "approximate": True,
+                    "status": "editing",
+                }
+            ],
+        },
+    ))
+
+    payload = json.loads(mock_ws.send.await_args.args[0])
+    assert payload == {
+        "event": "file_edit",
+        "chat_id": "chat-1",
+        "edits": [
+            {
+                "version": 1,
+                "phase": "start",
+                "call_id": "call-1",
+                "tool": "write_file",
+                "path": "src/app.py",
+                "added": 12,
+                "deleted": 2,
+                "approximate": True,
+                "status": "editing",
+            }
+        ],
+    }
+
+
 @pytest.mark.asyncio
 async def test_send_progress_includes_agent_ui_blob() -> None:
     bus = MagicMock()
@@ -707,7 +757,7 @@ async def test_maybe_push_turn_run_wall_clock_skips_when_no_active_turn() -> Non
     channel = WebSocketChannel({"enabled": True, "allowFrom": ["*"]}, bus)
     mock_ws = AsyncMock()
     channel._attach(mock_ws, "chat-1")
-    from nanobot.utils import webui_turn_helpers as wth
+    from nanobot.session import webui_turns as wth
 
     wth._WEBSOCKET_TURN_WALL_STARTED_AT.clear()
     await channel._maybe_push_turn_run_wall_clock("chat-1")
@@ -720,7 +770,7 @@ async def test_maybe_push_turn_run_wall_clock_replays_running() -> None:
     channel = WebSocketChannel({"enabled": True, "allowFrom": ["*"]}, bus)
     mock_ws = AsyncMock()
     channel._attach(mock_ws, "chat-1")
-    from nanobot.utils import webui_turn_helpers as wth
+    from nanobot.session import webui_turns as wth
 
     wth._WEBSOCKET_TURN_WALL_STARTED_AT.clear()
     try:
@@ -758,6 +808,25 @@ async def test_send_session_updated_emits_session_updated_event() -> None:
     assert body == {"event": "session_updated", "chat_id": "chat-1"}
 
 
+@pytest.mark.asyncio
+async def test_send_session_updated_includes_scope_when_present() -> None:
+    bus = MagicMock()
+    channel = WebSocketChannel({"enabled": True, "allowFrom": ["*"]}, bus)
+    mock_ws = AsyncMock()
+    channel._attach(mock_ws, "chat-1")
+
+    await channel.send(OutboundMessage(
+        channel="websocket",
+        chat_id="chat-1",
+        content="",
+        metadata={"_session_updated": True, "_session_update_scope": "metadata"},
+    ))
+
+    mock_ws.send.assert_awaited_once()
+    body = json.loads(mock_ws.send.await_args.args[0])
+    assert body == {"event": "session_updated", "chat_id": "chat-1", "scope": "metadata"}
+
+
 @pytest.mark.asyncio
 async def test_send_non_connection_closed_exception_is_raised() -> None:
     bus = MagicMock()
@@ -923,6 +992,11 @@ async def test_settings_api_returns_safe_subset_and_updates_whitelist(
     config = Config()
     config.agents.defaults.model = "openai/gpt-4o"
     config.providers.openai.api_key = "secret-key"
+    config.model_presets["deep"] = ModelPresetConfig(
+        model="anthropic/claude-opus-4-5",
+        provider="anthropic",
+        reasoning_effort="high",
+    )
     config.tools.web.search.provider = "brave"
     config.tools.web.search.api_key = "brave-secret"
     save_config(config, config_path)
@@ -943,16 +1017,52 @@ async def test_settings_api_returns_safe_subset_and_updates_whitelist(
         body = settings.json()
         assert body["agent"]["model"] == "openai/gpt-4o"
         assert body["agent"]["provider"] == "openai"
+        assert body["agent"]["model_preset"] == "default"
+        assert body["agent"]["max_tokens"] == 8192
+        assert body["agent"]["timezone"] == "UTC"
+        assert body["agent"]["tool_hint_max_length"] == 40
+        presets = {preset["name"]: preset for preset in body["model_presets"]}
+        assert presets["default"]["active"] is True
+        assert presets["deep"]["reasoning_effort"] == "high"
         providers = {provider["name"]: provider for provider in body["providers"]}
         assert providers["openai"]["configured"] is True
         assert providers["openai"]["api_key_hint"] == "secr••••-key"
+        assert providers["azure_openai"]["api_key_required"] is True
         assert providers["openrouter"]["configured"] is False
+        assert providers["openrouter"]["api_key_required"] is True
+        assert providers["skywork"]["label"] == "Skywork"
+        assert providers["skywork"]["default_api_base"] == "https://api.apifree.ai/v1"
+        assert providers["ant_ling"]["label"] == "Ant Ling"
+        assert providers["ant_ling"]["default_api_base"] == "https://api.ant-ling.com/v1"
+        assert providers["atomic_chat"]["configured"] is False
+        assert providers["atomic_chat"]["api_key_required"] is False
+        assert providers["atomic_chat"]["default_api_base"] == "http://localhost:1337/v1"
         assert body["agent"]["has_api_key"] is True
         assert body["web_search"]["provider"] == "brave"
         assert body["web_search"]["api_key_hint"] == "brav••••cret"
+        assert body["web_search"]["max_results"] == 5
+        assert body["web"]["fetch"]["use_jina_reader"] is True
         search_providers = {provider["name"]: provider for provider in body["web_search"]["providers"]}
         assert search_providers["duckduckgo"]["credential"] == "none"
         assert search_providers["searxng"]["credential"] == "base_url"
+        assert body["image_generation"]["enabled"] is False
+        assert body["image_generation"]["provider"] == "openrouter"
+        assert body["image_generation"]["provider_configured"] is False
+        assert body["image_generation"]["default_aspect_ratio"] == "1:1"
+        image_providers = {
+            provider["name"]: provider
+            for provider in body["image_generation"]["providers"]
+        }
+        assert image_providers["openrouter"]["label"] == "OpenRouter"
+        assert image_providers["openrouter"]["configured"] is False
+        assert image_providers["gemini"]["label"] == "Gemini"
+        assert body["runtime"]["config_path"] == str(config_path)
+        workspace_path = body["runtime"]["workspace_path"].replace("\\", "/")
+        assert workspace_path.endswith("/.nanobot/workspace")
+        assert body["runtime"]["gateway_port"] == 18790
+        assert body["advanced"]["exec_enabled"] is True
+        assert body["advanced"]["mcp_server_count"] == 0
+        assert body["restart_required_sections"] == []
         assert "secret-key" not in settings.text
         assert "brave-secret" not in settings.text
 
@@ -967,38 +1077,137 @@ async def test_settings_api_returns_safe_subset_and_updates_whitelist(
         assert provider_body["requires_restart"] is False
         provider_rows = {provider["name"]: provider for provider in provider_body["providers"]}
         assert provider_rows["openrouter"]["configured"] is True
+        assert provider_body["image_generation"]["provider_configured"] is True
         assert "sk-or-test" not in provider_updated.text
 
+        local_provider_updated = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/provider/update?provider=atomic_chat"
+            "&api_base=http%3A%2F%2Flocalhost%3A1337%2Fv1",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert local_provider_updated.status_code == 200
+        local_provider_body = local_provider_updated.json()
+        local_provider_rows = {
+            provider["name"]: provider for provider in local_provider_body["providers"]
+        }
+        assert local_provider_rows["atomic_chat"]["configured"] is True
+        assert "localhost:1337" in local_provider_updated.text
+
         updated = await _http_get(
             "http://127.0.0.1:"
-            f"{port}/api/settings/update?model=openrouter/test"
-            "&provider=openrouter",
+            f"{port}/api/settings/update?model=atomic_chat/test"
+            "&provider=atomic_chat&timezone=Asia%2FShanghai"
+            "&bot_name=Nano&bot_icon=N&tool_hint_max_length=120",
             headers={"Authorization": "Bearer tok"},
         )
         assert updated.status_code == 200
-        assert updated.json()["requires_restart"] is False
+        updated_body = updated.json()
+        assert updated_body["requires_restart"] is True
+        assert updated_body["restart_required_sections"] == ["runtime"]
+
+        preset_updated = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/update?model_preset=deep",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert preset_updated.status_code == 200
+        assert preset_updated.json()["agent"]["model"] == "anthropic/claude-opus-4-5"
+
+        bad_preset = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/update?model_preset=missing",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert bad_preset.status_code == 400
 
         search_updated = await _http_get(
             "http://127.0.0.1:"
             f"{port}/api/settings/web-search/update?provider=searxng"
-            "&base_url=https%3A%2F%2Fsearch.example.com",
+            "&base_url=https%3A%2F%2Fsearch.example.com"
+            "&max_results=8&timeout=45&use_jina_reader=false",
             headers={"Authorization": "Bearer tok"},
         )
         assert search_updated.status_code == 200
         search_body = search_updated.json()
-        assert search_body["requires_restart"] is False
+        assert search_body["requires_restart"] is True
+        assert search_body["restart_required_sections"] == ["runtime", "web"]
         assert search_body["web_search"]["provider"] == "searxng"
         assert search_body["web_search"]["api_key_hint"] is None
         assert search_body["web_search"]["base_url"] == "https://search.example.com"
+        assert search_body["web_search"]["max_results"] == 8
+        assert search_body["web"]["fetch"]["use_jina_reader"] is False
+
+        image_updated = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/image-generation/update?enabled=true"
+            "&provider=openrouter&model=openai%2Fgpt-image-1"
+            "&default_aspect_ratio=16%3A9&default_image_size=2K"
+            "&max_images_per_turn=3",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert image_updated.status_code == 200
+        image_body = image_updated.json()
+        assert image_body["requires_restart"] is True
+        assert image_body["restart_required_sections"] == ["image", "runtime", "web"]
+        assert image_body["image_generation"]["enabled"] is True
+        assert image_body["image_generation"]["model"] == "openai/gpt-image-1"
+        assert image_body["image_generation"]["default_aspect_ratio"] == "16:9"
+        assert image_body["image_generation"]["default_image_size"] == "2K"
+        assert image_body["image_generation"]["max_images_per_turn"] == 3
+
+        image_provider_updated = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/provider/update?provider=openrouter"
+            "&api_key=sk-or-next&api_base=https%3A%2F%2Fopenrouter.ai%2Fapi%2Fv1",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert image_provider_updated.status_code == 200
+        assert image_provider_updated.json()["requires_restart"] is True
+        assert image_provider_updated.json()["restart_required_sections"] == [
+            "image",
+            "runtime",
+            "web",
+        ]
+        assert "sk-or-next" not in image_provider_updated.text
+
+        bad_web = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/web-search/update?provider=duckduckgo&max_results=99",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert bad_web.status_code == 400
+
+        bad_image = await _http_get(
+            "http://127.0.0.1:"
+            f"{port}/api/settings/image-generation/update?provider=missing",
+            headers={"Authorization": "Bearer tok"},
+        )
+        assert bad_image.status_code == 400
 
         saved = load_config(config_path)
-        assert saved.agents.defaults.model == "openrouter/test"
-        assert saved.agents.defaults.provider == "openrouter"
-        assert saved.providers.openrouter.api_key == "sk-or-test"
+        assert saved.agents.defaults.model == "atomic_chat/test"
+        assert saved.agents.defaults.provider == "atomic_chat"
+        assert saved.agents.defaults.model_preset == "deep"
+        assert saved.agents.defaults.timezone == "Asia/Shanghai"
+        assert saved.agents.defaults.bot_name == "Nano"
+        assert saved.agents.defaults.bot_icon == "N"
+        assert saved.agents.defaults.tool_hint_max_length == 120
+        assert saved.providers.openrouter.api_key == "sk-or-next"
         assert saved.providers.openrouter.api_base == "https://openrouter.ai/api/v1"
+        assert saved.providers.atomic_chat.api_base == "http://localhost:1337/v1"
         assert saved.tools.web.search.provider == "searxng"
         assert saved.tools.web.search.api_key == ""
         assert saved.tools.web.search.base_url == "https://search.example.com"
+        assert saved.tools.web.search.max_results == 8
+        assert saved.tools.web.search.timeout == 45
+        assert saved.tools.web.fetch.use_jina_reader is False
+        assert saved.tools.image_generation.enabled is True
+        assert saved.tools.image_generation.provider == "openrouter"
+        assert saved.tools.image_generation.model == "openai/gpt-image-1"
+        assert saved.tools.image_generation.default_aspect_ratio == "16:9"
+        assert saved.tools.image_generation.default_image_size == "2K"
+        assert saved.tools.image_generation.max_images_per_turn == 3
     finally:
         await channel.stop()
         await server_task
@@ -1043,7 +1252,7 @@ def test_settings_payload_normalizes_camel_case_provider(
     save_config(config, config_path)
     monkeypatch.setattr("nanobot.config.loader._current_config_path", config_path)
 
-    body = _ch(bus)._settings_payload()
+    body = settings_payload()
 
     assert body["agent"]["provider"] == "minimax_anthropic"
 
@@ -1460,6 +1669,54 @@ def test_parse_envelope_rejects_legacy_and_garbage() -> None:
     assert _parse_envelope('{"type":123}') is None
 
 
+def test_sessions_list_includes_active_run_started_at() -> None:
+    from websockets.datastructures import Headers
+    from websockets.http11 import Request
+
+    from nanobot.session import webui_turns as wth
+
+    bus = MagicMock()
+    channel = _ch(bus)
+    channel._api_tokens["tok"] = time.monotonic() + 300.0
+    channel._session_manager = MagicMock()
+    channel._session_manager.list_sessions.return_value = [
+        {
+            "key": "websocket:chat-1",
+            "created_at": "2026-05-19T10:00:00Z",
+            "updated_at": "2026-05-19T10:01:00Z",
+            "title": "Running",
+            "preview": "work",
+            "path": "/private/path",
+        },
+        {
+            "key": "cli:chat-2",
+            "created_at": "2026-05-19T10:00:00Z",
+            "updated_at": "2026-05-19T10:01:00Z",
+        },
+    ]
+
+    wth._WEBSOCKET_TURN_WALL_STARTED_AT.clear()
+    try:
+        wth._WEBSOCKET_TURN_WALL_STARTED_AT["chat-1"] = 1_700_000_000.0
+        req = Request("/api/sessions", Headers([("Authorization", "Bearer tok")]))
+        resp = channel._handle_sessions_list(req)
+    finally:
+        wth._WEBSOCKET_TURN_WALL_STARTED_AT.clear()
+
+    assert resp.status_code == 200
+    body = json.loads(resp.body.decode())
+    assert body["sessions"] == [
+        {
+            "key": "websocket:chat-1",
+            "created_at": "2026-05-19T10:00:00Z",
+            "updated_at": "2026-05-19T10:01:00Z",
+            "title": "Running",
+            "preview": "work",
+            "run_started_at": 1_700_000_000.0,
+        }
+    ]
+
+
 @pytest.mark.parametrize(
     ("value", "expected"),
     [
@@ -1486,7 +1743,7 @@ def test_handle_webui_thread_get_returns_json(tmp_path, monkeypatch) -> None:
     from websockets.datastructures import Headers
     from websockets.http11 import Request
 
-    from nanobot.utils.webui_transcript import append_transcript_object
+    from nanobot.webui.transcript import append_transcript_object
 
     monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
     key = "websocket:c1"

tests/channels/test_websocket_http_routes.py

@@ -6,6 +6,7 @@ import json
 from pathlib import Path
 from typing import Any
 from unittest.mock import AsyncMock, MagicMock
+from urllib.parse import urlencode
 
 import httpx
 import pytest
@@ -176,13 +177,62 @@ async def test_sessions_list_only_returns_websocket_sessions_by_default(
         await server_task
 
 
+@pytest.mark.asyncio
+async def test_webui_sidebar_state_routes_are_config_dir_scoped(
+    bus: MagicMock, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    sm = _seed_session(tmp_path, key="websocket:sidebar")
+    channel = _ch(bus, session_manager=sm, port=29911)
+    server_task = asyncio.create_task(channel.start())
+    await asyncio.sleep(0.3)
+    try:
+        boot = await _http_get("http://127.0.0.1:29911/webui/bootstrap")
+        token = boot.json()["token"]
+        auth = {"Authorization": f"Bearer {token}"}
+
+        initial = await _http_get(
+            "http://127.0.0.1:29911/api/webui/sidebar-state",
+            headers=auth,
+        )
+        assert initial.status_code == 200
+        assert initial.json()["schema_version"] == 1
+        assert initial.json()["pinned_keys"] == []
+
+        payload = {
+            "pinned_keys": ["websocket:sidebar"],
+            "archived_keys": ["websocket:old"],
+            "title_overrides": {"websocket:sidebar": "Pinned work"},
+            "view": {"density": "compact", "show_archived": True},
+        }
+        query = urlencode({"state": json.dumps(payload)})
+        updated = await _http_get(
+            f"http://127.0.0.1:29911/api/webui/sidebar-state/update?{query}",
+            headers=auth,
+        )
+        assert updated.status_code == 200
+        body = updated.json()
+        assert body["pinned_keys"] == ["websocket:sidebar"]
+        assert body["title_overrides"] == {"websocket:sidebar": "Pinned work"}
+        assert body["view"]["density"] == "compact"
+
+        state_path = tmp_path / "webui" / "sidebar-state.json"
+        assert state_path.is_file()
+        assert json.loads(state_path.read_text(encoding="utf-8"))["pinned_keys"] == [
+            "websocket:sidebar"
+        ]
+    finally:
+        await channel.stop()
+        await server_task
+
+
 @pytest.mark.asyncio
 async def test_session_delete_removes_file(
     bus: MagicMock, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
 ) -> None:
     monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
     sm = _seed_session(tmp_path, key="websocket:doomed")
-    from nanobot.utils.webui_transcript import append_transcript_object
+    from nanobot.webui.transcript import append_transcript_object
 
     append_transcript_object("websocket:doomed", {"event": "user", "chat_id": "doomed", "text": "x"})
     channel = _ch(bus, session_manager=sm, port=29903)

tests/cli/test_commands.py

@@ -1170,6 +1170,7 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
             self.model = "test-model"
             self.provider = kwargs.get("provider", object())
             self.tools = {}
+            seen["agent"] = self
 
         async def process_direct(self, *_args, **_kwargs):
             return OutboundMessage(
@@ -1218,6 +1219,11 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
     assert isinstance(cron, _FakeCron)
     assert cron.on_job is not None
 
+    runtime_provider = object()
+    agent = seen["agent"]
+    agent.provider = runtime_provider
+    agent.model = "runtime-model"
+
     job = CronJob(
         id="cron-1",
         name="stretch",
@@ -1233,8 +1239,8 @@ def test_gateway_cron_evaluator_receives_scheduled_reminder_context(
 
     assert response == "Time to stretch."
     assert seen["response"] == "Time to stretch."
-    assert seen["provider"] is provider
-    assert seen["model"] == "test-model"
+    assert seen["provider"] is runtime_provider
+    assert seen["model"] == "runtime-model"
     assert seen["task_context"] == (
         "The scheduled time has arrived. Deliver this reminder to the user now, "
         "as a brief and natural message in their language. Speak directly to them — "
@@ -1543,6 +1549,9 @@ def test_gateway_health_endpoint_binds_and_serves_expected_responses(
             self.dream = _FakeDream()
             self.sessions = _FakeSessionManager()
 
+        def llm_runtime(self) -> None:
+            return None
+
         async def run(self) -> None:
             await asyncio.Event().wait()
 

tests/cli/test_interactive_retry_wait.py

@@ -69,6 +69,72 @@ async def test_reasoning_delta_displayed_when_show_reasoning_enabled():
     assert calls == ["I should search first."]
 
 
+@pytest.mark.asyncio
+async def test_reasoning_delta_buffers_until_sentence_boundary():
+    calls: list[str] = []
+    channels_config = SimpleNamespace(
+        send_progress=True, send_tool_hints=False, show_reasoning=True,
+    )
+    reasoning_buffer = commands._ReasoningBuffer()
+
+    with patch("nanobot.cli.commands._print_cli_reasoning", side_effect=lambda t, th, r=None: calls.append(t)):
+        first = await commands._maybe_print_interactive_progress(
+            SimpleNamespace(
+                content="The",
+                metadata={"_progress": True, "_reasoning_delta": True},
+            ),
+            None,
+            channels_config,
+            reasoning_buffer=reasoning_buffer,
+        )
+        second = await commands._maybe_print_interactive_progress(
+            SimpleNamespace(
+                content=" user asked.",
+                metadata={"_progress": True, "_reasoning_delta": True},
+            ),
+            None,
+            channels_config,
+            reasoning_buffer=reasoning_buffer,
+        )
+
+    assert first is True
+    assert second is True
+    assert calls == ["The user asked."]
+
+
+@pytest.mark.asyncio
+async def test_reasoning_end_flushes_buffered_delta():
+    calls: list[str] = []
+    channels_config = SimpleNamespace(
+        send_progress=True, send_tool_hints=False, show_reasoning=True,
+    )
+    reasoning_buffer = commands._ReasoningBuffer()
+
+    with patch("nanobot.cli.commands._print_cli_reasoning", side_effect=lambda t, th, r=None: calls.append(t)):
+        delta = await commands._maybe_print_interactive_progress(
+            SimpleNamespace(
+                content="The user asked",
+                metadata={"_progress": True, "_reasoning_delta": True},
+            ),
+            None,
+            channels_config,
+            reasoning_buffer=reasoning_buffer,
+        )
+        end = await commands._maybe_print_interactive_progress(
+            SimpleNamespace(
+                content="",
+                metadata={"_progress": True, "_reasoning_end": True},
+            ),
+            None,
+            channels_config,
+            reasoning_buffer=reasoning_buffer,
+        )
+
+    assert delta is True
+    assert end is True
+    assert calls == ["The user asked"]
+
+
 @pytest.mark.asyncio
 async def test_reasoning_hidden_when_show_reasoning_disabled():
     """Reasoning content should be suppressed when show_reasoning is False."""

tests/providers/test_ant_ling_provider.py

@@ -0,0 +1,73 @@
+"""Tests for the Ant Ling provider registration."""
+
+from unittest.mock import patch
+
+from nanobot.config.schema import Config, ProvidersConfig
+from nanobot.providers.openai_compat_provider import OpenAICompatProvider
+from nanobot.providers.registry import PROVIDERS, find_by_name
+
+
+def test_ant_ling_config_field_exists() -> None:
+    config = ProvidersConfig()
+
+    assert hasattr(config, "ant_ling")
+
+
+def test_ant_ling_provider_in_registry() -> None:
+    specs = {spec.name: spec for spec in PROVIDERS}
+
+    assert "ant_ling" in specs
+    ant_ling = specs["ant_ling"]
+    assert ant_ling.backend == "openai_compat"
+    assert ant_ling.env_key == "ANT_LING_API_KEY"
+    assert ant_ling.display_name == "Ant Ling"
+    assert ant_ling.default_api_base == "https://api.ant-ling.com/v1"
+
+
+def test_find_by_name_accepts_ant_ling_spellings() -> None:
+    spec = find_by_name("ant_ling")
+
+    assert spec is not None
+    assert find_by_name("ant-ling") is spec
+    assert find_by_name("antLing") is spec
+
+
+def test_ant_ling_model_auto_matches_with_default_api_base() -> None:
+    config = Config.model_validate({
+        "providers": {
+            "antLing": {
+                "apiKey": "ling-key",
+            },
+        },
+        "agents": {
+            "defaults": {
+                "model": "Ling-2.6-flash",
+            },
+        },
+    })
+
+    assert config.get_provider_name("Ling-2.6-flash") == "ant_ling"
+    assert config.get_api_key("Ling-2.6-flash") == "ling-key"
+    assert config.get_api_base("Ling-2.6-flash") == "https://api.ant-ling.com/v1"
+
+
+def test_ant_ling_preserves_official_model_name() -> None:
+    spec = find_by_name("ant_ling")
+    with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI"):
+        provider = OpenAICompatProvider(
+            api_key="ling-key",
+            default_model="Ling-2.6-flash",
+            spec=spec,
+        )
+
+    kwargs = provider._build_kwargs(
+        messages=[{"role": "user", "content": "hi"}],
+        tools=None,
+        model="Ling-2.6-flash",
+        max_tokens=1024,
+        temperature=0.7,
+        reasoning_effort=None,
+        tool_choice=None,
+    )
+
+    assert kwargs["model"] == "Ling-2.6-flash"

tests/providers/test_anthropic_stream_idle.py

@@ -129,6 +129,74 @@ async def test_chat_stream_invokes_on_thinking_delta_for_thinking_delta() -> Non
     assert text_parts == ["X"]
 
 
+@pytest.mark.asyncio
+async def test_chat_stream_invokes_tool_call_delta_for_input_json_delta() -> None:
+    provider = AnthropicProvider(api_key="sk-test")
+    provider._client = MagicMock()
+
+    chunks = [
+        SimpleNamespace(
+            type="content_block_start",
+            index=1,
+            content_block=SimpleNamespace(
+                type="tool_use",
+                id="toolu_1",
+                name="write_file",
+            ),
+        ),
+        SimpleNamespace(
+            type="content_block_delta",
+            index=1,
+            delta=SimpleNamespace(
+                type="input_json_delta",
+                partial_json='{"path":"notes.md","content":"',
+            ),
+        ),
+        SimpleNamespace(
+            type="content_block_delta",
+            index=1,
+            delta=SimpleNamespace(type="input_json_delta", partial_json="line\\n"),
+        ),
+    ]
+    fake = _FakeAsyncStream(chunks)
+    stream_cm = MagicMock()
+    stream_cm.__aenter__ = AsyncMock(return_value=fake)
+    stream_cm.__aexit__ = AsyncMock(return_value=None)
+    provider._client.messages.stream = MagicMock(return_value=stream_cm)
+
+    deltas: list[dict] = []
+
+    async def on_tool_delta(delta: dict) -> None:
+        deltas.append(delta)
+
+    await provider.chat_stream(
+        messages=[{"role": "user", "content": "write"}],
+        on_tool_call_delta=on_tool_delta,
+    )
+
+    assert deltas == [
+        {
+            "index": 1,
+            "call_id": "toolu_1",
+            "name": "write_file",
+            "arguments_delta": "",
+        },
+        {
+            "index": 1,
+            "call_id": "toolu_1",
+            "name": "write_file",
+            "arguments_delta": '{"path":"notes.md","content":"',
+        },
+        {
+            "index": 1,
+            "call_id": "toolu_1",
+            "name": "write_file",
+            "arguments_delta": "line\\n",
+        },
+    ]
+    fake.get_final_message.assert_awaited_once()
+
+
 @pytest.mark.asyncio
 async def test_chat_stream_without_callback_still_finalizes() -> None:
     provider = AnthropicProvider(api_key="sk-test")

tests/providers/test_image_generation.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import base64
 from pathlib import Path
 from typing import Any
 
@@ -8,9 +9,12 @@ import pytest
 
 from nanobot.providers.image_generation import (
     AIHubMixImageGenerationClient,
+    GeminiImageGenerationClient,
     GeneratedImageResponse,
     ImageGenerationError,
+    MiniMaxImageGenerationClient,
     OpenRouterImageGenerationClient,
+    StepFunImageGenerationClient,
 )
 
 PNG_BYTES = (
@@ -23,6 +27,7 @@ PNG_DATA_URL = (
     "data:image/png;base64,"
     "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mP8/x8AAwMCAO+/p9sAAAAASUVORK5CYII="
 )
+JPEG_BYTES = b"\xff\xd8\xff\xe0" + b"0" * 12
 
 
 class FakeResponse:
@@ -202,3 +207,311 @@ async def test_aihubmix_image_generation_downloads_url_response() -> None:
 
     assert response.images[0].startswith("data:image/png;base64,")
     assert fake.get_calls[0]["url"] == "https://cdn.example/image.png"
+
+
+@pytest.mark.asyncio
+async def test_aihubmix_base64_response_uses_detected_mime() -> None:
+    raw_b64 = base64.b64encode(JPEG_BYTES).decode("ascii")
+    fake = FakeClient(FakeResponse({"output": {"b64_json": raw_b64}}))
+    client = AIHubMixImageGenerationClient(
+        api_key="sk-ahm-test",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    response = await client.generate(prompt="draw", model="gpt-image-2-free")
+
+    assert response.images == [f"data:image/jpeg;base64,{raw_b64}"]
+
+
+RAW_B64 = PNG_DATA_URL.removeprefix("data:image/png;base64,")
+
+
+@pytest.mark.asyncio
+async def test_gemini_imagen_payload_and_response() -> None:
+    fake = FakeClient(
+        FakeResponse({"predictions": [{"bytesBase64Encoded": RAW_B64, "mimeType": "image/png"}]})
+    )
+    client = GeminiImageGenerationClient(
+        api_key="AIza-test",
+        api_base="https://generativelanguage.googleapis.com/v1beta",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    response = await client.generate(
+        prompt="a sunset",
+        model="imagen-4.0-generate-001",
+        aspect_ratio="16:9",
+    )
+
+    assert response.images == [PNG_DATA_URL]
+    assert response.content == ""
+    call = fake.calls[0]
+    assert call["url"].endswith(":predict")
+    assert call["headers"]["x-goog-api-key"] == "AIza-test"
+    assert "params" not in call
+    body = call["json"]
+    assert body["instances"] == [{"prompt": "a sunset"}]
+    assert body["parameters"]["sampleCount"] == 1
+    assert body["parameters"]["aspectRatio"] == "16:9"
+
+
+@pytest.mark.asyncio
+async def test_gemini_imagen_ignores_unsupported_aspect_ratio() -> None:
+    fake = FakeClient(
+        FakeResponse({"predictions": [{"bytesBase64Encoded": RAW_B64, "mimeType": "image/png"}]})
+    )
+    client = GeminiImageGenerationClient(api_key="AIza-test", client=fake)  # type: ignore[arg-type]
+
+    await client.generate(prompt="a sunset", model="imagen-4.0-generate-001", aspect_ratio="2:3")
+
+    body = fake.calls[0]["json"]
+    assert "aspectRatio" not in body["parameters"]
+
+
+@pytest.mark.asyncio
+async def test_gemini_flash_payload_and_response() -> None:
+    fake = FakeClient(
+        FakeResponse(
+            {
+                "candidates": [
+                    {
+                        "content": {
+                            "parts": [
+                                {"text": "here is your image"},
+                                {"inlineData": {"mimeType": "image/png", "data": RAW_B64}},
+                            ]
+                        }
+                    }
+                ]
+            }
+        )
+    )
+    client = GeminiImageGenerationClient(
+        api_key="AIza-test",
+        api_base="https://generativelanguage.googleapis.com/v1beta",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    response = await client.generate(
+        prompt="draw a cat",
+        model="gemini-2.0-flash-preview-image-generation",
+    )
+
+    assert response.images == [PNG_DATA_URL]
+    assert response.content == "here is your image"
+    call = fake.calls[0]
+    assert call["url"].endswith(":generateContent")
+    assert call["headers"]["x-goog-api-key"] == "AIza-test"
+    assert "params" not in call
+    body = call["json"]
+    assert body["generationConfig"]["responseModalities"] == ["TEXT", "IMAGE"]
+    assert body["contents"][0]["parts"][-1] == {"text": "draw a cat"}
+
+
+@pytest.mark.asyncio
+async def test_gemini_flash_reference_images(tmp_path: Path) -> None:
+    ref = tmp_path / "ref.png"
+    ref.write_bytes(PNG_BYTES)
+    fake = FakeClient(
+        FakeResponse(
+            {
+                "candidates": [
+                    {
+                        "content": {
+                            "parts": [{"inlineData": {"mimeType": "image/png", "data": RAW_B64}}]
+                        }
+                    }
+                ]
+            }
+        )
+    )
+    client = GeminiImageGenerationClient(api_key="AIza-test", client=fake)  # type: ignore[arg-type]
+
+    response = await client.generate(
+        prompt="edit this",
+        model="gemini-2.0-flash-preview-image-generation",
+        reference_images=[str(ref)],
+    )
+
+    assert response.images == [PNG_DATA_URL]
+    parts = fake.calls[0]["json"]["contents"][0]["parts"]
+    assert parts[0]["inlineData"]["mimeType"] == "image/png"
+    assert parts[0]["inlineData"]["data"].startswith("iVBOR")
+    assert parts[1] == {"text": "edit this"}
+
+
+@pytest.mark.asyncio
+async def test_gemini_requires_api_key() -> None:
+    client = GeminiImageGenerationClient(api_key=None)
+
+    with pytest.raises(ImageGenerationError, match="API key"):
+        await client.generate(prompt="draw", model="imagen-4.0-generate-001")
+
+
+@pytest.mark.asyncio
+async def test_gemini_no_images_raises() -> None:
+    fake = FakeClient(FakeResponse({"candidates": [{"content": {"parts": [{"text": "sorry"}]}}]}))
+    client = GeminiImageGenerationClient(api_key="AIza-test", client=fake)  # type: ignore[arg-type]
+
+    with pytest.raises(ImageGenerationError, match="returned no images"):
+        await client.generate(prompt="draw", model="gemini-2.0-flash-preview-image-generation")
+
+
+@pytest.mark.asyncio
+async def test_minimax_payload_and_response_with_reference_image(tmp_path: Path) -> None:
+    ref = tmp_path / "ref.png"
+    ref.write_bytes(PNG_BYTES)
+    fake = FakeClient(FakeResponse({"data": {"image_base64": [RAW_B64]}}))
+    client = MiniMaxImageGenerationClient(
+        api_key="sk-mm-test",
+        api_base="https://api.minimaxi.com/v1/",
+        extra_headers={"X-Test": "1"},
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    response = await client.generate(
+        prompt="draw a character",
+        model="image-01",
+        reference_images=[str(ref)],
+        aspect_ratio="21:9",
+    )
+
+    assert response.images == [PNG_DATA_URL]
+    call = fake.calls[0]
+    assert call["url"] == "https://api.minimaxi.com/v1/image_generation"
+    assert call["headers"]["Authorization"] == "Bearer sk-mm-test"
+    assert call["headers"]["X-Test"] == "1"
+    body = call["json"]
+    assert body["model"] == "image-01"
+    assert body["prompt"] == "draw a character"
+    assert body["response_format"] == "base64"
+    assert body["aspect_ratio"] == "21:9"
+    assert body["subject_reference"][0]["type"] == "character"
+    assert body["subject_reference"][0]["image_file"].startswith("data:image/png;base64,")
+
+
+# ---------------------------------------------------------------------------
+# StepFun (阶跃星辰)
+# ---------------------------------------------------------------------------
+
+@pytest.mark.asyncio
+async def test_stepfun_payload_and_response_with_aspect_ratio() -> None:
+    fake = FakeClient(FakeResponse({"data": [{"b64_json": RAW_B64}]}))
+    client = StepFunImageGenerationClient(
+        api_key="sk-sf-test",
+        api_base="https://api.stepfun.com/v1",
+        extra_headers={"X-Test": "1"},
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    response = await client.generate(
+        prompt="a cat on the moon",
+        model="step-image-edit-2",
+        aspect_ratio="16:9",
+    )
+
+    assert response.images == [PNG_DATA_URL]
+    call = fake.calls[0]
+    assert call["url"] == "https://api.stepfun.com/v1/images/generations"
+    assert call["headers"]["Authorization"] == "Bearer sk-sf-test"
+    assert call["headers"]["X-Test"] == "1"
+    body = call["json"]
+    assert body["model"] == "step-image-edit-2"
+    assert body["prompt"] == "a cat on the moon"
+    assert body["response_format"] == "b64_json"
+    assert body["n"] == 1
+    assert body["size"] == "1280x800"
+
+
+@pytest.mark.asyncio
+async def test_stepfun_default_size_when_no_aspect_ratio() -> None:
+    fake = FakeClient(FakeResponse({"data": [{"b64_json": RAW_B64}]}))
+    client = StepFunImageGenerationClient(
+        api_key="sk-sf-test",
+        api_base="https://api.stepfun.com/v1",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    await client.generate(prompt="a dog", model="step-image-edit-2")
+
+    body = fake.calls[0]["json"]
+    assert body["size"] == "1024x1024"
+
+
+@pytest.mark.asyncio
+async def test_stepfun_uses_explicit_image_size() -> None:
+    fake = FakeClient(FakeResponse({"data": [{"b64_json": RAW_B64}]}))
+    client = StepFunImageGenerationClient(
+        api_key="sk-sf-test",
+        api_base="https://api.stepfun.com/v1",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    await client.generate(
+        prompt="a bird",
+        model="step-image-edit-2",
+        image_size="1024x1024",
+    )
+
+    body = fake.calls[0]["json"]
+    assert body["size"] == "1024x1024"
+
+
+@pytest.mark.asyncio
+async def test_stepfun_style_reference_on_1x_model(tmp_path: Path) -> None:
+    """step-1x-medium supports style_reference for reference-image generation."""
+    ref = tmp_path / "ref.png"
+    ref.write_bytes(PNG_BYTES)
+    fake = FakeClient(FakeResponse({"data": [{"b64_json": RAW_B64}]}))
+    client = StepFunImageGenerationClient(
+        api_key="sk-sf-test",
+        api_base="https://api.stepfun.com/v1",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    await client.generate(
+        prompt="in this style",
+        model="step-1x-medium",
+        reference_images=[str(ref)],
+    )
+
+    body = fake.calls[0]["json"]
+    assert "style_reference" in body
+    assert body["style_reference"]["source_url"].startswith("data:image/png;base64,")
+
+
+@pytest.mark.asyncio
+async def test_stepfun_no_style_reference_on_non_1x_model() -> None:
+    """step-image-edit-2 does not use style_reference; reference images are ignored."""
+    fake = FakeClient(FakeResponse({"data": [{"b64_json": RAW_B64}]}))
+    client = StepFunImageGenerationClient(
+        api_key="sk-sf-test",
+        api_base="https://api.stepfun.com/v1",
+        client=fake,  # type: ignore[arg-type]
+    )
+
+    await client.generate(
+        prompt="a flower",
+        model="step-image-edit-2",
+        reference_images=["/tmp/ref.png"],
+    )
+
+    body = fake.calls[0]["json"]
+    assert "style_reference" not in body
+
+
+@pytest.mark.asyncio
+async def test_stepfun_requires_api_key() -> None:
+    client = StepFunImageGenerationClient(api_key=None)
+
+    with pytest.raises(ImageGenerationError, match="API key"):
+        await client.generate(prompt="draw", model="step-image-edit-2")
+
+
+@pytest.mark.asyncio
+async def test_stepfun_no_images_raises() -> None:
+    fake = FakeClient(FakeResponse({"data": [{"text": "sorry"}]}))
+    client = StepFunImageGenerationClient(api_key="sk-sf-test", client=fake)  # type: ignore[arg-type]
+
+    with pytest.raises(ImageGenerationError, match="returned no images"):
+        await client.generate(prompt="draw", model="step-image-edit-2")

tests/providers/test_litellm_kwargs.py

@@ -164,6 +164,130 @@ def _fake_chat_stream_reasoning_chunks():
     return _stream()
 
 
+def _fake_chat_stream_tool_call_chunks():
+    """Mimic OpenAI-compatible streaming tool-call argument deltas."""
+
+    async def _stream():
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason=None,
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=[
+                            SimpleNamespace(
+                                index=0,
+                                id="call_write",
+                                function=SimpleNamespace(
+                                    name="write_file",
+                                    arguments='{"path":"notes.md","content":"',
+                                ),
+                            )
+                        ],
+                    ),
+                ),
+            ],
+            usage=None,
+        )
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason=None,
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=[
+                            SimpleNamespace(
+                                index=0,
+                                id=None,
+                                function=SimpleNamespace(name=None, arguments='line\\n"}'),
+                            )
+                        ],
+                    ),
+                ),
+            ],
+            usage=None,
+        )
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason="tool_calls",
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=None,
+                    ),
+                ),
+            ],
+            usage=SimpleNamespace(prompt_tokens=10, completion_tokens=5, total_tokens=15),
+        )
+
+    return _stream()
+
+
+def _fake_chat_stream_legacy_function_call_chunks():
+    """Mimic older OpenAI-compatible ``delta.function_call`` chunks."""
+
+    async def _stream():
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason=None,
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=None,
+                        function_call=SimpleNamespace(
+                            name="write_file",
+                            arguments='{"path":"notes.md","content":"',
+                        ),
+                    ),
+                ),
+            ],
+            usage=None,
+        )
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason=None,
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=None,
+                        function_call=SimpleNamespace(
+                            name=None,
+                            arguments='line\\n"}',
+                        ),
+                    ),
+                ),
+            ],
+            usage=None,
+        )
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    finish_reason="function_call",
+                    delta=SimpleNamespace(
+                        content=None,
+                        reasoning_content=None,
+                        reasoning=None,
+                        tool_calls=None,
+                        function_call=None,
+                    ),
+                ),
+            ],
+            usage=SimpleNamespace(prompt_tokens=10, completion_tokens=5, total_tokens=15),
+        )
+
+    return _stream()
+
+
 @pytest.mark.asyncio
 async def test_openai_compat_stream_forwards_reasoning_deltas_deepseek_style() -> None:
     """Regression: DeepSeek-V4 / reasoner expose ``delta.reasoning_content`` during streaming."""
@@ -202,6 +326,98 @@ async def test_openai_compat_stream_forwards_reasoning_deltas_deepseek_style() -
     mock_chat.assert_awaited_once()
 
 
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    ("provider_name", "model"),
+    [
+        ("openai", "gpt-4o"),
+        ("deepseek", "deepseek-chat"),
+        ("minimax", "MiniMax-M2.7"),
+        ("zhipu", "glm-4.6"),
+    ],
+)
+async def test_openai_compat_stream_forwards_tool_call_argument_deltas(
+    provider_name: str,
+    model: str,
+) -> None:
+    mock_chat = AsyncMock(return_value=_fake_chat_stream_tool_call_chunks())
+    spec = find_by_name(provider_name)
+    deltas: list[dict] = []
+
+    async def on_tool_delta(delta: dict) -> None:
+        deltas.append(delta)
+
+    with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI") as mock_openai:
+        client_instance = mock_openai.return_value
+        client_instance.chat.completions.create = mock_chat
+
+        provider = OpenAICompatProvider(
+            api_key="sk-test",
+            default_model=model,
+            spec=spec,
+        )
+        result = await provider.chat_stream(
+            messages=[{"role": "user", "content": "write"}],
+            tools=[{"type": "function", "function": {"name": "write_file"}}],
+            model=model,
+            on_tool_call_delta=on_tool_delta,
+        )
+
+    assert deltas == [
+        {
+            "index": 0,
+            "call_id": "call_write",
+            "name": "write_file",
+            "arguments_delta": '{"path":"notes.md","content":"',
+        },
+        {"index": 0, "call_id": "", "name": "", "arguments_delta": 'line\\n"}'},
+    ]
+    assert result.tool_calls[0].name == "write_file"
+    assert result.tool_calls[0].arguments == {"path": "notes.md", "content": "line\n"}
+    kwargs = mock_chat.await_args.kwargs
+    if provider_name == "zhipu":
+        assert kwargs["extra_body"]["tool_stream"] is True
+    else:
+        assert kwargs.get("extra_body", {}).get("tool_stream") is None
+
+
+@pytest.mark.asyncio
+async def test_openai_compat_stream_forwards_legacy_function_call_argument_deltas() -> None:
+    mock_chat = AsyncMock(return_value=_fake_chat_stream_legacy_function_call_chunks())
+    deltas: list[dict] = []
+
+    async def on_tool_delta(delta: dict) -> None:
+        deltas.append(delta)
+
+    with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI") as mock_openai:
+        client_instance = mock_openai.return_value
+        client_instance.chat.completions.create = mock_chat
+
+        provider = OpenAICompatProvider(
+            api_key="sk-test",
+            default_model="deepseek-chat",
+            spec=find_by_name("deepseek"),
+        )
+        result = await provider.chat_stream(
+            messages=[{"role": "user", "content": "write"}],
+            tools=[{"type": "function", "function": {"name": "write_file"}}],
+            model="deepseek-chat",
+            on_tool_call_delta=on_tool_delta,
+        )
+
+    assert deltas == [
+        {
+            "index": 0,
+            "call_id": "",
+            "name": "write_file",
+            "arguments_delta": '{"path":"notes.md","content":"',
+        },
+        {"index": 0, "call_id": "", "name": "", "arguments_delta": 'line\\n"}'},
+    ]
+    assert result.tool_calls[0].name == "write_file"
+    assert result.tool_calls[0].arguments == {"path": "notes.md", "content": "line\n"}
+
+
 class _FakeResponsesError(Exception):
     def __init__(self, status_code: int, text: str):
         super().__init__(text)

tests/providers/test_llm_response.py

@@ -44,9 +44,15 @@ class TestShouldExecuteTools:
         resp = _response("stop")
         assert resp.should_execute_tools is True
 
+    def test_legacy_function_call_reason_executes(self) -> None:
+        # Older OpenAI-compatible streaming APIs can still use the singular
+        # function_call finish reason while carrying a tool-call-shaped payload.
+        resp = _response("function_call")
+        assert resp.should_execute_tools is True
+
     @pytest.mark.parametrize(
         "anomalous_reason",
-        ["refusal", "content_filter", "error", "length", "function_call", ""],
+        ["refusal", "content_filter", "error", "length", ""],
     )
     def test_tool_calls_under_anomalous_reason_blocked(self, anomalous_reason: str) -> None:
         # This is the #3220 bug: gateways injecting tool_calls under any of these

tests/providers/test_openai_codex_provider.py

@@ -16,7 +16,15 @@ async def test_codex_prompt_cache_key_uses_stable_conversation_prefix(monkeypatc
         lambda: SimpleNamespace(account_id="acct", access="token"),
     )
 
-    async def fake_request(url, headers, body, verify, on_content_delta=None):
+    async def fake_request(
+        url,
+        headers,
+        body,
+        verify,
+        on_content_delta=None,
+        on_tool_call_delta=None,
+    ):
+        _ = on_tool_call_delta
         bodies.append(body)
         return "ok", [], "stop"
 

tests/providers/test_openai_responses.py

@@ -453,6 +453,56 @@ class TestConsumeSdkStream:
         assert tool_calls[0].name == "get_weather"
         assert tool_calls[0].arguments == {"city": "SF"}
 
+    @pytest.mark.asyncio
+    async def test_tool_call_argument_delta_callback(self):
+        item_added = MagicMock(type="function_call", call_id="c1", id="fc1", arguments="")
+        item_added.name = "write_file"
+        ev1 = MagicMock(type="response.output_item.added", item=item_added)
+        ev2 = MagicMock(
+            type="response.function_call_arguments.delta",
+            call_id="c1",
+            delta='{"path":"a.txt","content":"',
+        )
+        ev3 = MagicMock(
+            type="response.function_call_arguments.delta",
+            call_id="c1",
+            delta='hello\\n',
+        )
+        ev4 = MagicMock(
+            type="response.function_call_arguments.done",
+            call_id="c1",
+            arguments='{"path":"a.txt","content":"hello\\n"}',
+        )
+        item_done = MagicMock(
+            type="function_call",
+            call_id="c1",
+            id="fc1",
+            arguments='{"path":"a.txt","content":"hello\\n"}',
+        )
+        item_done.name = "write_file"
+        ev5 = MagicMock(type="response.output_item.done", item=item_done)
+        resp_obj = MagicMock(status="completed", usage=None, output=[])
+        ev6 = MagicMock(type="response.completed", response=resp_obj)
+        deltas: list[dict] = []
+
+        async def cb(delta: dict) -> None:
+            deltas.append(delta)
+
+        async def stream():
+            for e in [ev1, ev2, ev3, ev4, ev5, ev6]:
+                yield e
+
+        await consume_sdk_stream(stream(), on_tool_call_delta=cb)
+        assert deltas == [
+            {"call_id": "c1", "name": "write_file", "arguments_delta": ""},
+            {
+                "call_id": "c1",
+                "name": "write_file",
+                "arguments_delta": '{"path":"a.txt","content":"',
+            },
+            {"call_id": "c1", "name": "write_file", "arguments_delta": "hello\\n"},
+        ]
+
     @pytest.mark.asyncio
     async def test_usage_extracted(self):
         usage_obj = MagicMock(input_tokens=10, output_tokens=5, total_tokens=15)

tests/providers/test_provider_retry.py

@@ -242,7 +242,7 @@ async def test_image_fallback_returns_error_on_second_failure() -> None:
 
 @pytest.mark.asyncio
 async def test_image_fallback_without_meta_uses_default_placeholder() -> None:
-    """When _meta is absent, fallback placeholder is '[image omitted]'."""
+    """When _meta is absent, fallback placeholder is '[image]'."""
     provider = ScriptedProvider([
         LLMResponse(content="error", finish_reason="error"),
         LLMResponse(content="ok"),
@@ -256,7 +256,7 @@ async def test_image_fallback_without_meta_uses_default_placeholder() -> None:
     for msg in msgs_on_retry:
         content = msg.get("content")
         if isinstance(content, list):
-            assert any("[image omitted]" in (b.get("text") or "") for b in content)
+            assert any("[image]" in (b.get("text") or "") for b in content)
 
 
 @pytest.mark.asyncio

tests/providers/test_skywork_provider.py

@@ -0,0 +1,80 @@
+"""Tests for the Skywork provider registration."""
+
+from unittest.mock import patch
+
+from nanobot.config.schema import Config, ProvidersConfig
+from nanobot.providers.openai_compat_provider import OpenAICompatProvider
+from nanobot.providers.registry import PROVIDERS, find_by_name
+
+
+def test_skywork_config_field_exists() -> None:
+    config = ProvidersConfig()
+
+    assert hasattr(config, "skywork")
+
+
+def test_skywork_provider_in_registry() -> None:
+    specs = {spec.name: spec for spec in PROVIDERS}
+
+    assert "skywork" in specs
+    skywork = specs["skywork"]
+    assert skywork.backend == "openai_compat"
+    assert skywork.env_key == "SKYWORK_API_KEY"
+    assert ("APIFREE_API_KEY", "{api_key}") in skywork.env_extras
+    assert skywork.display_name == "Skywork"
+    assert skywork.is_gateway is True
+    assert skywork.detect_by_base_keyword == "apifree.ai"
+    assert skywork.default_api_base == "https://api.apifree.ai/v1"
+    assert skywork.supports_max_completion_tokens is False
+
+
+def test_find_by_name_skywork() -> None:
+    spec = find_by_name("skywork")
+
+    assert spec is not None
+    assert spec.name == "skywork"
+
+
+def test_skywork_model_auto_matches_with_default_api_base() -> None:
+    config = Config.model_validate(
+        {
+            "providers": {
+                "skywork": {
+                    "apiKey": "sky-key",
+                },
+            },
+            "agents": {
+                "defaults": {
+                    "model": "skywork-ai/skyclaw-v1",
+                },
+            },
+        }
+    )
+
+    assert config.get_provider_name("skywork-ai/skyclaw-v1") == "skywork"
+    assert config.get_api_key("skywork-ai/skyclaw-v1") == "sky-key"
+    assert config.get_api_base("skywork-ai/skyclaw-v1") == "https://api.apifree.ai/v1"
+
+
+def test_skywork_preserves_model_id_and_uses_chat_completion_max_tokens() -> None:
+    spec = find_by_name("skywork")
+    with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI"):
+        provider = OpenAICompatProvider(
+            api_key="sky-key",
+            default_model="skywork-ai/skyclaw-v1",
+            spec=spec,
+        )
+
+    kwargs = provider._build_kwargs(
+        messages=[{"role": "user", "content": "hi"}],
+        tools=None,
+        model="skywork-ai/skyclaw-v1",
+        max_tokens=1024,
+        temperature=0.7,
+        reasoning_effort=None,
+        tool_choice=None,
+    )
+
+    assert kwargs["model"] == "skywork-ai/skyclaw-v1"
+    assert kwargs["max_tokens"] == 1024
+    assert "max_completion_tokens" not in kwargs

tests/test_truncate_text_shadowing.py

@@ -29,3 +29,47 @@ def test_sanitize_persisted_blocks_truncate_text_shadowing_regression() -> None:
     assert isinstance(out[0]["text"], str)
     assert out[0]["text"] != content[0]["text"]
 
+
+def test_sanitize_persisted_blocks_strips_audio_and_video() -> None:
+    """Audio and video blocks with base64 payloads must be replaced with placeholders."""
+    from nanobot.agent.loop import AgentLoop
+
+    dummy = SimpleNamespace(max_tool_result_chars=1000)
+    content = [
+        {"type": "text", "text": "analyze this"},
+        {
+            "type": "input_audio",
+            "input_audio": {"data": "aGVsbG8=", "format": "wav"},
+            "_meta": {"path": "/tmp/voice.wav"},
+        },
+        {
+            "type": "video_url",
+            "video_url": {"url": "data:video/mp4;base64,aGVsbG8="},
+            "_meta": {"path": "/tmp/clip.mp4"},
+        },
+    ]
+
+    out = AgentLoop._sanitize_persisted_blocks(dummy, content)
+
+    assert len(out) == 3
+    assert out[0] == content[0]
+    assert out[1] == {"type": "text", "text": "[audio: /tmp/voice.wav]"}
+    assert out[2] == {"type": "text", "text": "[video: /tmp/clip.mp4]"}
+
+
+def test_sanitize_persisted_blocks_strips_audio_video_without_meta() -> None:
+    """When _meta is absent, fallback placeholders use bare label."""
+    from nanobot.agent.loop import AgentLoop
+
+    dummy = SimpleNamespace(max_tool_result_chars=1000)
+    content = [
+        {"type": "input_audio", "input_audio": {"data": "aGVsbG8=", "format": "wav"}},
+        {"type": "video_url", "video_url": {"url": "data:video/mp4;base64,aGVsbG8="}},
+    ]
+
+    out = AgentLoop._sanitize_persisted_blocks(dummy, content)
+
+    assert len(out) == 2
+    assert out[0] == {"type": "text", "text": "[audio]"}
+    assert out[1] == {"type": "text", "text": "[video]"}
+

tests/tools/test_image_generation_tool.py

@@ -44,8 +44,8 @@ async def test_generate_image_tool_stores_artifact_and_source_images(
     set_config_path(tmp_path / "config.json")
     FakeImageClient.instances = []
     monkeypatch.setattr(
-        "nanobot.agent.tools.image_generation.OpenRouterImageGenerationClient",
-        FakeImageClient,
+        "nanobot.agent.tools.image_generation.get_image_gen_provider",
+        lambda name: FakeImageClient if name == "openrouter" else None,
     )
     ref = tmp_path / "ref.png"
     ref.write_bytes(PNG_BYTES)
@@ -98,8 +98,8 @@ async def test_generate_image_tool_selects_aihubmix_provider(
     set_config_path(tmp_path / "config.json")
     FakeImageClient.instances = []
     monkeypatch.setattr(
-        "nanobot.agent.tools.image_generation.AIHubMixImageGenerationClient",
-        FakeImageClient,
+        "nanobot.agent.tools.image_generation.get_image_gen_provider",
+        lambda name: FakeImageClient if name == "aihubmix" else None,
     )
     tool = ImageGenerationTool(
         workspace=tmp_path,

tests/utils/test_artifacts.py

@@ -10,8 +10,6 @@ from nanobot.config.loader import set_config_path
 from nanobot.utils.artifacts import (
     ArtifactError,
     decode_image_data_url,
-    generated_image_paths_from_messages,
-    generated_image_tool_result,
     store_generated_image_artifact,
 )
 
@@ -66,22 +64,3 @@ def test_store_generated_image_artifact_rejects_unsafe_save_dir(tmp_path: Path)
             model="m",
             save_dir="../outside",
         )
-
-
-def test_generated_image_paths_from_tool_results() -> None:
-    result = generated_image_tool_result(
-        [
-            {"id": "img_1", "path": "/tmp/one.png"},
-            {"id": "img_2", "path": "/tmp/two.png"},
-        ]
-    )
-    payload = json.loads(result)
-
-    assert generated_image_paths_from_messages(
-        [
-            {"role": "tool", "name": "generate_image", "content": result},
-            {"role": "tool", "name": "other", "content": result},
-        ]
-    ) == ["/tmp/one.png", "/tmp/two.png"]
-    assert "runtime attaches generated images automatically" in payload["next_step"]
-    assert "Do not call message" in payload["next_step"]

tests/utils/test_file_edit_events.py

@@ -0,0 +1,392 @@
+from __future__ import annotations
+
+import asyncio
+from pathlib import Path
+from types import SimpleNamespace
+
+from nanobot.utils.file_edit_events import (
+    build_file_edit_end_event,
+    build_file_edit_start_event,
+    line_diff_stats,
+    prepare_file_edit_tracker,
+    read_file_snapshot,
+    StreamingFileEditTracker,
+)
+
+
+def test_line_diff_stats_counts_replacements_insertions_and_deletions() -> None:
+    added, deleted = line_diff_stats("a\nb\nc\n", "a\nB\nc\nd\n")
+    assert (added, deleted) == (2, 1)
+
+
+def test_line_diff_stats_normalizes_crlf() -> None:
+    assert line_diff_stats("a\r\nb\r\n", "a\nb\nc\n") == (1, 0)
+
+
+def test_line_diff_stats_counts_new_file_crlf_lines_once() -> None:
+    assert line_diff_stats("", "a\r\nb\r\n") == (2, 0)
+
+
+def test_write_file_start_predicts_and_end_calibrates_exact_diff(tmp_path: Path) -> None:
+    target = tmp_path / "notes.txt"
+    target.write_text("old\nkeep\n", encoding="utf-8")
+    params = {"path": "notes.txt", "content": "new\nkeep\nextra\n"}
+    tracker = prepare_file_edit_tracker(
+        call_id="call-write",
+        tool_name="write_file",
+        tool=None,
+        workspace=tmp_path,
+        params=params,
+    )
+
+    assert tracker is not None
+    start = build_file_edit_start_event(tracker, params)
+    assert start == {
+        "version": 1,
+        "call_id": "call-write",
+        "tool": "write_file",
+        "path": "notes.txt",
+        "absolute_path": (tmp_path / "notes.txt").resolve().as_posix(),
+        "phase": "start",
+        "added": 2,
+        "deleted": 1,
+        "approximate": True,
+        "status": "editing",
+    }
+
+    target.write_text("new\nkeep\nextra\n", encoding="utf-8")
+    end = build_file_edit_end_event(tracker)
+    assert end["phase"] == "end"
+    assert end["status"] == "done"
+    assert end["approximate"] is False
+    assert (end["added"], end["deleted"]) == (2, 1)
+
+
+def test_binary_file_is_reported_but_not_counted(tmp_path: Path) -> None:
+    target = tmp_path / "data.bin"
+    target.write_bytes(b"\x00\x01before")
+    tracker = prepare_file_edit_tracker(
+        call_id="call-bin",
+        tool_name="edit_file",
+        tool=None,
+        workspace=tmp_path,
+        params={"path": "data.bin", "old_text": "before", "new_text": "after"},
+    )
+
+    assert tracker is not None
+    assert not read_file_snapshot(target).countable
+    target.write_bytes(b"\x00\x01after")
+    event = build_file_edit_end_event(tracker)
+    assert event["binary"] is True
+    assert (event["added"], event["deleted"]) == (0, 0)
+
+
+def test_oversized_write_file_end_uses_known_content_for_exact_count(tmp_path: Path) -> None:
+    target = tmp_path / "large.txt"
+    params = {"path": "large.txt", "content": "x" * (2 * 1024 * 1024 + 1)}
+    tracker = prepare_file_edit_tracker(
+        call_id="call-large",
+        tool_name="write_file",
+        tool=None,
+        workspace=tmp_path,
+        params=params,
+    )
+
+    assert tracker is not None
+    target.write_text(params["content"], encoding="utf-8")
+    event = build_file_edit_end_event(tracker, params)
+    assert event.get("binary") is not True
+    assert event["added"] == 1
+    assert event["deleted"] == 0
+
+
+def test_streaming_write_file_tracker_emits_live_line_counts(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"path":"notes.md","content":"',
+        })
+        await tracker.update({
+            "index": 0,
+            "arguments_delta": "line\\n" * 24,
+        })
+
+    asyncio.run(run())
+
+    assert events[0] == {
+        "version": 1,
+        "call_id": "call-live",
+        "tool": "write_file",
+        "path": "notes.md",
+        "absolute_path": (tmp_path / "notes.md").resolve().as_posix(),
+        "phase": "start",
+        "added": 0,
+        "deleted": 0,
+        "approximate": True,
+        "status": "editing",
+    }
+    assert events[-1]["path"] == "notes.md"
+    assert events[-1]["status"] == "editing"
+    assert events[-1]["approximate"] is True
+    assert events[-1]["added"] == 24
+    assert events[-1]["deleted"] == 0
+
+
+def test_streaming_write_file_tracker_emits_pending_before_path(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"content":"line\\n',
+        })
+        await tracker.update({
+            "index": 0,
+            "arguments_delta": 'more\\n","path":"late.md"',
+        })
+
+    asyncio.run(run())
+
+    assert events[0] == {
+        "version": 1,
+        "call_id": "call-live",
+        "tool": "write_file",
+        "path": "",
+        "phase": "start",
+        "added": 1,
+        "deleted": 0,
+        "approximate": True,
+        "status": "editing",
+        "pending": True,
+    }
+    assert events[-1]["path"] == "late.md"
+    assert events[-1].get("pending") is not True
+    assert events[-1]["added"] == 2
+
+
+def test_streaming_write_file_tracker_flushes_small_pending_count(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"path":"small.md","content":"one\\n',
+        })
+        await tracker.flush()
+
+    asyncio.run(run())
+    assert events
+    assert events[-1]["path"] == "small.md"
+    assert events[-1]["added"] == 1
+
+
+def test_streaming_write_file_tracker_normalizes_crlf_line_counts(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"path":"windows.txt","content":"one\\r\\ntwo\\r\\n',
+        })
+        await tracker.flush()
+
+    asyncio.run(run())
+    assert events[-1]["path"] == "windows.txt"
+    assert events[-1]["added"] == 2
+
+
+def test_streaming_write_file_tracker_counts_unicode_escaped_newlines(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"path":"unicode.txt","content":"one\\u000atwo',
+        })
+        await tracker.flush()
+
+    asyncio.run(run())
+    assert events[-1]["path"] == "unicode.txt"
+    assert events[-1]["added"] == 2
+
+
+def test_streaming_edit_file_tracker_emits_live_line_counts(tmp_path: Path) -> None:
+    target = tmp_path / "notes.md"
+    target.write_text("old\nkeep\n", encoding="utf-8")
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-edit",
+            "name": "edit_file",
+            "arguments_delta": '{"path":"notes.md","old_text":"old\\nkeep","new_text":"',
+        })
+        await tracker.update({
+            "index": 0,
+            "arguments_delta": "new\\nkeep\\nextra\\n" * 8,
+        })
+
+    asyncio.run(run())
+
+    assert events[0] == {
+        "version": 1,
+        "call_id": "call-edit",
+        "tool": "edit_file",
+        "path": "notes.md",
+        "absolute_path": (tmp_path / "notes.md").resolve().as_posix(),
+        "phase": "start",
+        "added": 0,
+        "deleted": 2,
+        "approximate": True,
+        "status": "editing",
+    }
+    assert events[-1]["path"] == "notes.md"
+    assert events[-1]["status"] == "editing"
+    assert events[-1]["approximate"] is True
+    assert events[-1]["added"] == 24
+    assert events[-1]["deleted"] == 2
+
+
+def test_streaming_tracker_applies_canonical_call_id_to_final_tool(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "name": "write_file",
+            "arguments_delta": '{"path":"matched.md","content":"one\\n',
+        })
+        final = SimpleNamespace(
+            id="provider-final-id",
+            name="write_file",
+            arguments={"path": "matched.md", "content": "one\n"},
+        )
+        tracker.apply_final_call_ids([final])
+        assert final.id == "idx:0"
+
+    asyncio.run(run())
+
+
+def test_streaming_edit_file_tracker_flushes_small_pending_count(tmp_path: Path) -> None:
+    target = tmp_path / "small.py"
+    target.write_text("old\n", encoding="utf-8")
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-edit",
+            "name": "edit_file",
+            "arguments_delta": '{"path":"small.py","old_text":"old\\n","new_text":"new\\nextra',
+        })
+        await tracker.flush()
+
+    asyncio.run(run())
+    assert events
+    assert events[-1]["path"] == "small.py"
+    assert events[-1]["added"] == 2
+    assert events[-1]["deleted"] == 1
+
+
+def test_streaming_write_file_tracker_errors_unmatched_live_edits(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "call-live",
+            "name": "write_file",
+            "arguments_delta": '{"path":"aborted.md","content":"one\\n',
+        })
+        await tracker.error_unmatched([], "Tool call did not complete.")
+
+    asyncio.run(run())
+    assert events[-1]["path"] == "aborted.md"
+    assert events[-1]["phase"] == "error"
+    assert events[-1]["status"] == "error"
+
+
+def test_streaming_write_file_tracker_keeps_matched_final_tool_call(tmp_path: Path) -> None:
+    events: list[dict] = []
+
+    async def emit(batch: list[dict]) -> None:
+        events.extend(batch)
+
+    async def run() -> None:
+        tracker = StreamingFileEditTracker(workspace=tmp_path, tools={}, emit=emit)
+        await tracker.update({
+            "index": 0,
+            "call_id": "idx-only",
+            "name": "write_file",
+            "arguments_delta": '{"path":"matched.md","content":"one\\n',
+        })
+        await tracker.error_unmatched([
+            SimpleNamespace(
+                id="final-call",
+                name="write_file",
+                arguments={"path": "matched.md", "content": "one\n"},
+            )
+        ], "Tool call did not complete.")
+
+    asyncio.run(run())
+    assert events
+    assert all(event["status"] == "editing" for event in events)
+
+
+def test_untracked_tools_do_not_prepare_file_edit_tracker(tmp_path: Path) -> None:
+    assert prepare_file_edit_tracker(
+        call_id="call-exec",
+        tool_name="exec",
+        tool=None,
+        workspace=tmp_path,
+        params={"path": "created-by-shell.txt"},
+    ) is None

tests/utils/test_webui_compat_imports.py

@@ -0,0 +1,14 @@
+import importlib
+
+from nanobot.session import webui_turns
+from nanobot.webui import thread_disk, transcript
+
+
+def test_legacy_webui_utils_imports_resolve_to_new_modules() -> None:
+    legacy_thread_disk = importlib.import_module("nanobot.utils.webui_thread_disk")
+    legacy_transcript = importlib.import_module("nanobot.utils.webui_transcript")
+    legacy_turn_helpers = importlib.import_module("nanobot.utils.webui_turn_helpers")
+
+    assert legacy_thread_disk.delete_webui_thread is thread_disk.delete_webui_thread
+    assert legacy_transcript.append_transcript_object is transcript.append_transcript_object
+    assert legacy_turn_helpers.mark_webui_session is webui_turns.mark_webui_session

tests/utils/test_webui_sidebar_state.py

@@ -0,0 +1,73 @@
+import json
+
+from nanobot.webui.sidebar_state import (
+    default_webui_sidebar_state,
+    read_webui_sidebar_state,
+    webui_sidebar_state_path,
+    write_webui_sidebar_state,
+)
+
+
+def test_sidebar_state_defaults_when_file_missing(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+
+    state = read_webui_sidebar_state()
+
+    assert state == default_webui_sidebar_state()
+    assert webui_sidebar_state_path() == tmp_path / "webui" / "sidebar-state.json"
+
+
+def test_sidebar_state_normalizes_old_or_partial_payload(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    path = webui_sidebar_state_path()
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(
+        json.dumps(
+            {
+                "pinned_keys": ["websocket:a", "websocket:a", "", 123],
+                "archived_keys": ["websocket:b"],
+                "title_overrides": {"websocket:a": "  Release notes  ", "bad": ""},
+                "tags_by_key": {"websocket:a": ["work", "work", ""]},
+                "collapsed_groups": {"Earlier": 1},
+                "view": {"density": "tiny", "show_archived": True, "sort": "nope"},
+            }
+        ),
+        encoding="utf-8",
+    )
+
+    state = read_webui_sidebar_state()
+
+    assert state["schema_version"] == 1
+    assert state["pinned_keys"] == ["websocket:a"]
+    assert state["archived_keys"] == ["websocket:b"]
+    assert state["title_overrides"] == {"websocket:a": "Release notes"}
+    assert state["tags_by_key"] == {"websocket:a": ["work"]}
+    assert state["collapsed_groups"] == {"Earlier": True}
+    assert state["view"] == {
+        "density": "comfortable",
+        "show_previews": False,
+        "show_timestamps": False,
+        "show_archived": True,
+        "sort": "updated_desc",
+    }
+
+
+def test_sidebar_state_write_is_scoped_to_config_data_dir(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+
+    state = write_webui_sidebar_state(
+        {
+            "pinned_keys": ["websocket:a"],
+            "archived_keys": ["websocket:b"],
+            "title_overrides": {"websocket:a": "Release"},
+            "view": {"density": "compact", "show_previews": True},
+        }
+    )
+
+    assert state["pinned_keys"] == ["websocket:a"]
+    assert state["archived_keys"] == ["websocket:b"]
+    assert state["title_overrides"] == {"websocket:a": "Release"}
+    assert state["view"]["density"] == "compact"
+    assert state["view"]["show_previews"] is True
+    assert webui_sidebar_state_path().is_file()
+    assert read_webui_sidebar_state()["pinned_keys"] == ["websocket:a"]

tests/utils/test_webui_thread_disk.py

@@ -2,8 +2,8 @@
 
 from __future__ import annotations
 
-from nanobot.utils.webui_thread_disk import delete_webui_thread, webui_thread_file_path
-from nanobot.utils.webui_transcript import append_transcript_object, webui_transcript_path
+from nanobot.webui.thread_disk import delete_webui_thread, webui_thread_file_path
+from nanobot.webui.transcript import append_transcript_object, webui_transcript_path
 
 
 def test_delete_webui_thread_removes_legacy_json_and_transcript(tmp_path, monkeypatch) -> None:

tests/utils/test_webui_transcript.py

@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from nanobot.utils.webui_transcript import (
+from nanobot.webui.transcript import (
     WEBUI_TRANSCRIPT_SCHEMA_VERSION,
     append_transcript_object,
     read_transcript_lines,
@@ -42,8 +42,306 @@ def test_replay_delta_and_turn_end(tmp_path, monkeypatch) -> None:
     assert msgs[1]["latencyMs"] == 42
 
 
+def test_replay_file_edit_event_creates_file_activity(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    key = "websocket:t-file"
+    for ev in (
+        {"event": "user", "chat_id": "t-file", "text": "edit"},
+        {
+            "event": "message",
+            "chat_id": "t-file",
+            "text": 'write_file({"path":"foo.txt"})',
+            "kind": "tool_hint",
+        },
+        {
+            "event": "file_edit",
+            "chat_id": "t-file",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-write",
+                    "tool": "write_file",
+                    "path": "foo.txt",
+                    "phase": "end",
+                    "added": 2,
+                    "deleted": 1,
+                    "approximate": False,
+                    "status": "done",
+                },
+            ],
+        },
+    ):
+        append_transcript_object(key, ev)
+
+    msgs = replay_transcript_to_ui_messages(read_transcript_lines(key))
+
+    assert len(msgs) == 3
+    assert msgs[1]["kind"] == "trace"
+    assert msgs[1]["traces"] == ['write_file({"path":"foo.txt"})']
+    assert "fileEdits" not in msgs[1]
+    assert msgs[2]["kind"] == "trace"
+    assert msgs[2]["traces"] == []
+    assert msgs[2]["fileEdits"] == [
+        {
+            "version": 1,
+            "call_id": "call-write",
+            "tool": "write_file",
+            "path": "foo.txt",
+            "phase": "end",
+            "added": 2,
+            "deleted": 1,
+            "approximate": False,
+            "status": "done",
+        },
+    ]
+    assert msgs[2]["activitySegmentId"]
+    assert msgs[2]["activitySegmentId"] != msgs[1]["activitySegmentId"]
+
+
+def test_replay_tool_events_dedupes_finish_after_start() -> None:
+    msgs = replay_transcript_to_ui_messages([
+        {
+            "event": "message",
+            "chat_id": "t-tool",
+            "text": 'exec({"cmd":"ls"})',
+            "kind": "tool_hint",
+            "tool_events": [
+                {
+                    "phase": "start",
+                    "call_id": "call-exec",
+                    "name": "exec",
+                    "arguments": {"cmd": "ls"},
+                },
+            ],
+        },
+        {
+            "event": "message",
+            "chat_id": "t-tool",
+            "text": "",
+            "kind": "progress",
+            "tool_events": [
+                {
+                    "phase": "end",
+                    "call_id": "call-exec",
+                    "name": "exec",
+                    "arguments": {"cmd": "ls"},
+                    "result": "ok",
+                },
+                {
+                    "phase": "end",
+                    "call_id": "call-read",
+                    "name": "read_file",
+                    "arguments": {"path": "notes.md"},
+                    "result": "done",
+                },
+            ],
+        },
+    ])
+
+    assert len(msgs) == 1
+    assert msgs[0]["traces"] == [
+        'exec({"cmd": "ls"})',
+        'read_file({"path": "notes.md"})',
+    ]
+
+
+def test_replay_file_edit_progress_merges_after_interleaved_activity(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    key = "websocket:t-file-progress"
+    for ev in (
+        {"event": "user", "chat_id": "t-file-progress", "text": "edit"},
+        {
+            "event": "message",
+            "chat_id": "t-file-progress",
+            "text": 'write_file({"path":"foo.txt"})',
+            "kind": "tool_hint",
+        },
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-progress",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-write",
+                    "tool": "write_file",
+                    "path": "foo.txt",
+                    "phase": "start",
+                    "added": 12,
+                    "deleted": 0,
+                    "approximate": True,
+                    "status": "editing",
+                },
+            ],
+        },
+        {
+            "event": "message",
+            "chat_id": "t-file-progress",
+            "text": "still working",
+            "kind": "progress",
+        },
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-progress",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-write",
+                    "tool": "write_file",
+                    "path": "foo.txt",
+                    "phase": "end",
+                    "added": 30,
+                    "deleted": 0,
+                    "approximate": False,
+                    "status": "done",
+                },
+            ],
+        },
+    ):
+        append_transcript_object(key, ev)
+
+    msgs = replay_transcript_to_ui_messages(read_transcript_lines(key))
+    file_edit_messages = [msg for msg in msgs if msg.get("fileEdits")]
+
+    assert len(file_edit_messages) == 1
+    assert file_edit_messages[0]["fileEdits"] == [
+        {
+            "version": 1,
+            "call_id": "call-write",
+            "tool": "write_file",
+            "path": "foo.txt",
+            "phase": "end",
+            "added": 30,
+            "deleted": 0,
+            "approximate": False,
+            "status": "done",
+        },
+    ]
+
+
+def test_replay_file_edit_pending_placeholder_upgrades_to_path(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    key = "websocket:t-file-pending"
+    for ev in (
+        {"event": "user", "chat_id": "t-file-pending", "text": "write"},
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-pending",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-write",
+                    "tool": "write_file",
+                    "path": "",
+                    "phase": "start",
+                    "added": 1,
+                    "deleted": 0,
+                    "approximate": True,
+                    "status": "editing",
+                    "pending": True,
+                },
+            ],
+        },
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-pending",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-write",
+                    "tool": "write_file",
+                    "path": "foo.txt",
+                    "phase": "start",
+                    "added": 12,
+                    "deleted": 0,
+                    "approximate": True,
+                    "status": "editing",
+                },
+            ],
+        },
+    ):
+        append_transcript_object(key, ev)
+
+    msgs = replay_transcript_to_ui_messages(read_transcript_lines(key))
+    file_edit_messages = [msg for msg in msgs if msg.get("fileEdits")]
+
+    assert len(file_edit_messages) == 1
+    assert file_edit_messages[0]["fileEdits"] == [
+        {
+            "version": 1,
+            "call_id": "call-write",
+            "tool": "write_file",
+            "path": "foo.txt",
+            "phase": "start",
+            "added": 12,
+            "deleted": 0,
+            "approximate": True,
+            "status": "editing",
+        },
+    ]
+
+
+def test_replay_keeps_new_file_edit_after_reasoning_in_order(tmp_path, monkeypatch) -> None:
+    monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
+    key = "websocket:t-file-order"
+    for ev in (
+        {"event": "user", "chat_id": "t-file-order", "text": "edit"},
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-order",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-one",
+                    "tool": "write_file",
+                    "path": "one.txt",
+                    "phase": "start",
+                    "added": 10,
+                    "deleted": 0,
+                    "approximate": True,
+                    "status": "editing",
+                },
+            ],
+        },
+        {"event": "reasoning_delta", "chat_id": "t-file-order", "text": "Check next."},
+        {"event": "reasoning_end", "chat_id": "t-file-order"},
+        {
+            "event": "file_edit",
+            "chat_id": "t-file-order",
+            "edits": [
+                {
+                    "version": 1,
+                    "call_id": "call-two",
+                    "tool": "write_file",
+                    "path": "two.txt",
+                    "phase": "start",
+                    "added": 20,
+                    "deleted": 0,
+                    "approximate": True,
+                    "status": "editing",
+                },
+            ],
+        },
+    ):
+        append_transcript_object(key, ev)
+
+    msgs = replay_transcript_to_ui_messages(read_transcript_lines(key))
+
+    assert [msg.get("fileEdits", [{}])[0].get("path") if msg.get("fileEdits") else msg.get("reasoning") for msg in msgs[1:]] == [
+        "one.txt",
+        "Check next.",
+        "two.txt",
+    ]
+    file_edit_segments = [
+        msg.get("activitySegmentId")
+        for msg in msgs
+        if msg.get("fileEdits")
+    ]
+    assert len(file_edit_segments) == 2
+    assert file_edit_segments[0] != file_edit_segments[1]
+
+
 def test_build_response_schema(monkeypatch, tmp_path) -> None:
-    from nanobot.utils.webui_transcript import build_webui_thread_response
+    from nanobot.webui.transcript import build_webui_thread_response
 
     monkeypatch.setattr("nanobot.config.paths.get_data_dir", lambda: tmp_path)
     key = "websocket:t3"

tests/utils/test_webui_turn_helpers.py

@@ -5,7 +5,7 @@ from unittest.mock import AsyncMock, MagicMock
 import pytest
 
 from nanobot.bus.events import InboundMessage
-from nanobot.utils import webui_turn_helpers as wth
+from nanobot.session import webui_turns as wth
 
 
 @pytest.fixture(autouse=True)

webui/README.md

@@ -8,15 +8,11 @@ on the same port.
 For the project overview, install guide, and general docs map, see the root
 [`README.md`](../README.md).
 
-## Current status
+## Just want to use the WebUI?
 
-> [!NOTE]
-> The standalone WebUI development workflow currently requires a source
-> checkout.
->
-> WebUI changes in the GitHub repository may land before they are included in
-> the next packaged release, so source installs and published package versions
-> are not yet guaranteed to move in lockstep.
+If you installed nanobot via `pip install nanobot-ai`, the WebUI is **already bundled** in the wheel. Enable the WebSocket channel in `~/.nanobot/config.json` and run `nanobot gateway` — see the root [`README.md`](../README.md#-webui) for the 3-step setup. You do **not** need anything in this directory.
+
+This `webui/` tree is for people **hacking on the WebUI itself** (UI changes, new components, styling, etc.).
 
 ## Layout
 
@@ -25,7 +21,7 @@ webui/                 source tree (this directory)
 nanobot/web/dist/      build output served by the gateway
 ```
 
-## Develop from source
+## Develop the WebUI (Vite HMR)
 
 ### 1. Install nanobot from source
 
@@ -35,6 +31,8 @@ From the repository root:
 pip install -e .
 ```
 
+> Editable installs intentionally **skip** the WebUI bundle step — Vite HMR is faster than rebuilding `dist/` on every change.
+
 ### 2. Enable the WebSocket channel
 
 In `~/.nanobot/config.json`:
@@ -63,8 +61,7 @@ bun run dev
 
 Then open `http://127.0.0.1:5173`.
 
-By default, the dev server proxies `/api`, `/webui`, `/auth`, and WebSocket
-traffic to `http://127.0.0.1:8765`.
+By default the dev server proxies `/api`, `/webui`, `/auth`, and WebSocket traffic to `http://127.0.0.1:8765`.
 
 If your gateway listens on a non-default port, point the dev server at it:
 
@@ -74,7 +71,7 @@ NANOBOT_API_URL=http://127.0.0.1:9000 bun run dev
 
 ### Access from another device (LAN)
 
-To use the webui from another device on the same network, set `host` to `"0.0.0.0"` and configure a `token` or `tokenIssueSecret` in `~/.nanobot/config.json`:
+To use the WebUI from another device on the same network, set `host` to `"0.0.0.0"` and configure a `token` or `tokenIssueSecret` in `~/.nanobot/config.json`:
 
 ```json
 {
@@ -91,20 +88,20 @@ To use the webui from another device on the same network, set `host` to `"0.0.0.
 
 The gateway will refuse to start if `host` is `"0.0.0.0"` and neither `token` nor `tokenIssueSecret` is set.
 
-Then open `http://<your-ip>:8765` on the other device. The webui will show an authentication form where you enter the secret. It is saved in your browser so you only need to enter it once.
+Then open `http://<your-ip>:8765` on the other device. The WebUI will show an authentication form where you enter the secret. It is saved in your browser so you only need to enter it once.
 
 ## Build for packaged runtime
 
+You usually do not need to run this by hand: `python -m build` invokes the WebUI build automatically when packaging the wheel.
+
+If you want to preview the production bundle locally without rebuilding the wheel:
+
 ```bash
 cd webui
-bun run build
+bun run build          # writes to ../nanobot/web/dist
 ```
 
-This writes the production assets to `../nanobot/web/dist`, which is the
-directory served by `nanobot gateway` and bundled into the Python wheel.
-
-If you are cutting a release, run the build before packaging so the published
-wheel contains the current WebUI assets.
+The gateway picks up the new bundle on the next restart.
 
 ## Test
 

webui/bun.lock

@@ -15,12 +15,15 @@
         "@radix-ui/react-tooltip": "^1.1.6",
         "class-variance-authority": "^0.7.1",
         "clsx": "^2.1.1",
+        "i18next": "^26.0.6",
         "lucide-react": "^0.469.0",
         "react": "^18.3.1",
         "react-dom": "^18.3.1",
+        "react-i18next": "^17.0.4",
         "react-markdown": "^9.0.1",
         "react-syntax-highlighter": "^15.6.1",
         "rehype-katex": "^7.0.1",
+        "remark-breaks": "^4.0.0",
         "remark-gfm": "^4.0.0",
         "remark-math": "^6.0.0",
         "tailwind-merge": "^2.6.0",
@@ -506,8 +509,12 @@
 
     "highlightjs-vue": ["highlightjs-vue@1.0.0", "", {}, "sha512-PDEfEF102G23vHmPhLyPboFCD+BkMGu+GuJe2d9/eH4FsCwvgBpnc9n0pGE+ffKdph38s6foEZiEjdgHdzp+IA=="],
 
+    "html-parse-stringify": ["html-parse-stringify@3.0.1", "", { "dependencies": { "void-elements": "3.1.0" } }, "sha512-KknJ50kTInJ7qIScF3jeaFRpMpE8/lfiTdzf/twXyPBLAGrLRTmkz3AdTnKeh40X8k9L2fdYwEp/42WGXIRGcg=="],
+
     "html-url-attributes": ["html-url-attributes@3.0.1", "", {}, "sha512-ol6UPyBWqsrO6EJySPz2O7ZSr856WDrEzM5zMqp+FJJLGMW35cLYmmZnl0vztAZxRUoNZJFTCohfjuIJ8I4QBQ=="],
 
+    "i18next": ["i18next@26.2.0", "", { "peerDependencies": { "typescript": "^5 || ^6" }, "optionalPeers": ["typescript"] }, "sha512-zwBHldHdTmwN7r6UNc7lC6GWNN+YYg3DrRSeHR5PRRBf5QnJZcYHrQc0uaU26qZeYxR7iFZD+Y315dPnKP47wA=="],
+
     "indent-string": ["indent-string@4.0.0", "", {}, "sha512-EdDDZu4A2OyIK7Lr/2zG+w5jmbuk1DVBnEwREQvBzspBJkCEbRa8GxU1lghYcaGJCnRWibjDXlq779X1/y5xwg=="],
 
     "inline-style-parser": ["inline-style-parser@0.2.7", "", {}, "sha512-Nb2ctOyNR8DqQoR0OwRG95uNWIC0C1lCgf5Naz5H6Ji72KZ8OcFZLz2P5sNgwlyoJ8Yif11oMuYs5pBQa86csA=="],
@@ -588,6 +595,8 @@
 
     "mdast-util-mdxjs-esm": ["mdast-util-mdxjs-esm@2.0.1", "", { "dependencies": { "@types/estree-jsx": "^1.0.0", "@types/hast": "^3.0.0", "@types/mdast": "^4.0.0", "devlop": "^1.0.0", "mdast-util-from-markdown": "^2.0.0", "mdast-util-to-markdown": "^2.0.0" } }, "sha512-EcmOpxsZ96CvlP03NghtH1EsLtr0n9Tm4lPUJUBccV9RwUOneqSycg19n5HGzCf+10LozMRSObtVr3ee1WoHtg=="],
 
+    "mdast-util-newline-to-break": ["mdast-util-newline-to-break@2.0.0", "", { "dependencies": { "@types/mdast": "^4.0.0", "mdast-util-find-and-replace": "^3.0.0" } }, "sha512-MbgeFca0hLYIEx/2zGsszCSEJJ1JSCdiY5xQxRcLDDGa8EPvlLPupJ4DSajbMPAnC0je8jfb9TiUATnxxrHUog=="],
+
     "mdast-util-phrasing": ["mdast-util-phrasing@4.1.0", "", { "dependencies": { "@types/mdast": "^4.0.0", "unist-util-is": "^6.0.0" } }, "sha512-TqICwyvJJpBwvGAMZjj4J2n0X8QWp21b9l0o7eXyVJ25YNWYbJDVIyD1bZXE6WtV6RmKJVYmQAKWa0zWOABz2w=="],
 
     "mdast-util-to-hast": ["mdast-util-to-hast@13.2.1", "", { "dependencies": { "@types/hast": "^3.0.0", "@types/mdast": "^4.0.0", "@ungap/structured-clone": "^1.0.0", "devlop": "^1.0.0", "micromark-util-sanitize-uri": "^2.0.0", "trim-lines": "^3.0.0", "unist-util-position": "^5.0.0", "unist-util-visit": "^5.0.0", "vfile": "^6.0.0" } }, "sha512-cctsq2wp5vTsLIcaymblUriiTcZd0CwWtCbLvrOzYCDZoWyMNV8sZ7krj09FSnsiJi3WVsHLM4k6Dq/yaPyCXA=="],
@@ -718,6 +727,8 @@
 
     "react-dom": ["react-dom@18.3.1", "", { "dependencies": { "loose-envify": "^1.1.0", "scheduler": "^0.23.2" }, "peerDependencies": { "react": "^18.3.1" } }, "sha512-5m4nQKp+rZRb09LNH59GM4BxTh9251/ylbKIbpe7TpGxfJ+9kv6BLkLBXIjjspbgbnIBNqlI23tRnTWT0snUIw=="],
 
+    "react-i18next": ["react-i18next@17.0.8", "", { "dependencies": { "@babel/runtime": "^7.29.2", "html-parse-stringify": "^3.0.1", "use-sync-external-store": "^1.6.0" }, "peerDependencies": { "i18next": ">= 26.2.0", "react": ">= 16.8.0", "typescript": "^5 || ^6" }, "optionalPeers": ["typescript"] }, "sha512-0ooKbGLU8JXhe1zwpQUWIeXSgLPOfwJmgheWRIUpcoA0CpyabpGhayjdG+/eA5esC1AQ8h2jWpXjJfzQzeDOCw=="],
+
     "react-is": ["react-is@17.0.2", "", {}, "sha512-w2GsyukL62IJnlaff/nRegPQR94C/XXamvMWmSHRJ4y7Ts/4ocGRmTHvOs8PSE6pB3dWOrD/nueuU5sduBsQ4w=="],
 
     "react-markdown": ["react-markdown@9.1.0", "", { "dependencies": { "@types/hast": "^3.0.0", "@types/mdast": "^4.0.0", "devlop": "^1.0.0", "hast-util-to-jsx-runtime": "^2.0.0", "html-url-attributes": "^3.0.0", "mdast-util-to-hast": "^13.0.0", "remark-parse": "^11.0.0", "remark-rehype": "^11.0.0", "unified": "^11.0.0", "unist-util-visit": "^5.0.0", "vfile": "^6.0.0" }, "peerDependencies": { "@types/react": ">=18", "react": ">=18" } }, "sha512-xaijuJB0kzGiUdG7nc2MOMDUDBWPyGAjZtUrow9XxUeua8IqeP+VlIfAZ3bphpcLTnSZXz6z9jcVC/TCwbfgdw=="],
@@ -742,6 +753,8 @@
 
     "rehype-katex": ["rehype-katex@7.0.1", "", { "dependencies": { "@types/hast": "^3.0.0", "@types/katex": "^0.16.0", "hast-util-from-html-isomorphic": "^2.0.0", "hast-util-to-text": "^4.0.0", "katex": "^0.16.0", "unist-util-visit-parents": "^6.0.0", "vfile": "^6.0.0" } }, "sha512-OiM2wrZ/wuhKkigASodFoo8wimG3H12LWQaH8qSPVJn9apWKFSH3YOCtbKpBorTVw/eI7cuT21XBbvwEswbIOA=="],
 
+    "remark-breaks": ["remark-breaks@4.0.0", "", { "dependencies": { "@types/mdast": "^4.0.0", "mdast-util-newline-to-break": "^2.0.0", "unified": "^11.0.0" } }, "sha512-IjEjJOkH4FuJvHZVIW0QCDWxcG96kCq7An/KVH2NfJe6rKZU2AsHeB3OEjPNRxi4QC34Xdx7I2KGYn6IpT7gxQ=="],
+
     "remark-gfm": ["remark-gfm@4.0.1", "", { "dependencies": { "@types/mdast": "^4.0.0", "mdast-util-gfm": "^3.0.0", "micromark-extension-gfm": "^3.0.0", "remark-parse": "^11.0.0", "remark-stringify": "^11.0.0", "unified": "^11.0.0" } }, "sha512-1quofZ2RQ9EWdeN34S79+KExV1764+wCUGop5CPL1WGdD0ocPpu91lzPGbwWMECpEpd42kJGQwzRfyov9j4yNg=="],
 
     "remark-math": ["remark-math@6.0.0", "", { "dependencies": { "@types/mdast": "^4.0.0", "mdast-util-math": "^3.0.0", "micromark-extension-math": "^3.0.0", "unified": "^11.0.0" } }, "sha512-MMqgnP74Igy+S3WwnhQ7kqGlEerTETXMvJhrUzDikVZ2/uogJCb+WHUg97hK9/jcfc0dkD73s3LN8zU49cTEtA=="],
@@ -860,6 +873,8 @@
 
     "vitest": ["vitest@2.1.9", "", { "dependencies": { "@vitest/expect": "2.1.9", "@vitest/mocker": "2.1.9", "@vitest/pretty-format": "^2.1.9", "@vitest/runner": "2.1.9", "@vitest/snapshot": "2.1.9", "@vitest/spy": "2.1.9", "@vitest/utils": "2.1.9", "chai": "^5.1.2", "debug": "^4.3.7", "expect-type": "^1.1.0", "magic-string": "^0.30.12", "pathe": "^1.1.2", "std-env": "^3.8.0", "tinybench": "^2.9.0", "tinyexec": "^0.3.1", "tinypool": "^1.0.1", "tinyrainbow": "^1.2.0", "vite": "^5.0.0", "vite-node": "2.1.9", "why-is-node-running": "^2.3.0" }, "peerDependencies": { "@edge-runtime/vm": "*", "@types/node": "^18.0.0 || >=20.0.0", "@vitest/browser": "2.1.9", "@vitest/ui": "2.1.9", "happy-dom": "*", "jsdom": "*" }, "optionalPeers": ["@edge-runtime/vm", "@types/node", "@vitest/browser", "@vitest/ui", "happy-dom", "jsdom"], "bin": { "vitest": "vitest.mjs" } }, "sha512-MSmPM9REYqDGBI8439mA4mWhV5sKmDlBKWIYbA3lRb2PTHACE0mgKwA8yQ2xq9vxDTuk4iPrECBAEW2aoFXY0Q=="],
 
+    "void-elements": ["void-elements@3.1.0", "", {}, "sha512-Dhxzh5HZuiHQhbvTW9AMetFfBHDMYpo23Uo9btPXgdYP+3T5S+p+jgNy7spra+veYhBP2dCSgxR/i2Y02h5/6w=="],
+
     "web-namespaces": ["web-namespaces@2.0.1", "", {}, "sha512-bKr1DkiNa2krS7qxNtdrtHAmzuYGFQLiQ13TsorsdT6ULTkPLKuu5+GsFpDlg6JFjUTwX2DyhMPG2be8uPrqsQ=="],
 
     "webidl-conversions": ["webidl-conversions@7.0.0", "", {}, "sha512-VwddBukDzu71offAQR975unBIGqfKZpM+8ZX6ySk8nYhVoo5CYaZyzt3YBvYtRtO+aoGlqxPg/B87NGVZ/fu6g=="],

webui/package-lock.json

@@ -26,6 +26,7 @@
         "react-markdown": "^9.0.1",
         "react-syntax-highlighter": "^15.6.1",
         "rehype-katex": "^7.0.1",
+        "remark-breaks": "^4.0.0",
         "remark-gfm": "^4.0.0",
         "remark-math": "^6.0.0",
         "tailwind-merge": "^2.6.0"
@@ -318,6 +319,278 @@
         "node": ">=6.9.0"
       }
     },
+    "node_modules/@esbuild/aix-ppc64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/aix-ppc64/-/aix-ppc64-0.21.5.tgz",
+      "integrity": "sha512-1SDgH6ZSPTlggy1yI6+Dbkiz8xzpHJEVAlF/AM1tHPLsf5STom9rwtjE4hKAF20FfXXNTFqEYXyJNWh1GiZedQ==",
+      "cpu": [
+        "ppc64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "aix"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/android-arm": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-arm/-/android-arm-0.21.5.tgz",
+      "integrity": "sha512-vCPvzSjpPHEi1siZdlvAlsPxXl7WbOVUBBAowWug4rJHb68Ox8KualB+1ocNvT5fjv6wpkX6o/iEpbDrf68zcg==",
+      "cpu": [
+        "arm"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/android-arm64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-arm64/-/android-arm64-0.21.5.tgz",
+      "integrity": "sha512-c0uX9VAUBQ7dTDCjq+wdyGLowMdtR/GoC2U5IYk/7D1H1JYC0qseD7+11iMP2mRLN9RcCMRcjC4YMclCzGwS/A==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/android-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/android-x64/-/android-x64-0.21.5.tgz",
+      "integrity": "sha512-D7aPRUUNHRBwHxzxRvp856rjUHRFW1SdQATKXH2hqA0kAZb1hKmi02OpYRacl0TxIGz/ZmXWlbZgjwWYaCakTA==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/darwin-arm64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/darwin-arm64/-/darwin-arm64-0.21.5.tgz",
+      "integrity": "sha512-DwqXqZyuk5AiWWf3UfLiRDJ5EDd49zg6O9wclZ7kUMv2WRFr4HKjXp/5t8JZ11QbQfUS6/cRCKGwYhtNAY88kQ==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/darwin-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/darwin-x64/-/darwin-x64-0.21.5.tgz",
+      "integrity": "sha512-se/JjF8NlmKVG4kNIuyWMV/22ZaerB+qaSi5MdrXtd6R08kvs2qCN4C09miupktDitvh8jRFflwGFBQcxZRjbw==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/freebsd-arm64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/freebsd-arm64/-/freebsd-arm64-0.21.5.tgz",
+      "integrity": "sha512-5JcRxxRDUJLX8JXp/wcBCy3pENnCgBR9bN6JsY4OmhfUtIHe3ZW0mawA7+RDAcMLrMIZaf03NlQiX9DGyB8h4g==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/freebsd-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/freebsd-x64/-/freebsd-x64-0.21.5.tgz",
+      "integrity": "sha512-J95kNBj1zkbMXtHVH29bBriQygMXqoVQOQYA+ISs0/2l3T9/kj42ow2mpqerRBxDJnmkUDCaQT/dfNXWX/ZZCQ==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-arm": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-arm/-/linux-arm-0.21.5.tgz",
+      "integrity": "sha512-bPb5AHZtbeNGjCKVZ9UGqGwo8EUu4cLq68E95A53KlxAPRmUyYv2D6F0uUI65XisGOL1hBP5mTronbgo+0bFcA==",
+      "cpu": [
+        "arm"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-arm64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-arm64/-/linux-arm64-0.21.5.tgz",
+      "integrity": "sha512-ibKvmyYzKsBeX8d8I7MH/TMfWDXBF3db4qM6sy+7re0YXya+K1cem3on9XgdT2EQGMu4hQyZhan7TeQ8XkGp4Q==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-ia32": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-ia32/-/linux-ia32-0.21.5.tgz",
+      "integrity": "sha512-YvjXDqLRqPDl2dvRODYmmhz4rPeVKYvppfGYKSNGdyZkA01046pLWyRKKI3ax8fbJoK5QbxblURkwK/MWY18Tg==",
+      "cpu": [
+        "ia32"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-loong64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-loong64/-/linux-loong64-0.21.5.tgz",
+      "integrity": "sha512-uHf1BmMG8qEvzdrzAqg2SIG/02+4/DHB6a9Kbya0XDvwDEKCoC8ZRWI5JJvNdUjtciBGFQ5PuBlpEOXQj+JQSg==",
+      "cpu": [
+        "loong64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-mips64el": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-mips64el/-/linux-mips64el-0.21.5.tgz",
+      "integrity": "sha512-IajOmO+KJK23bj52dFSNCMsz1QP1DqM6cwLUv3W1QwyxkyIWecfafnI555fvSGqEKwjMXVLokcV5ygHW5b3Jbg==",
+      "cpu": [
+        "mips64el"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-ppc64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-ppc64/-/linux-ppc64-0.21.5.tgz",
+      "integrity": "sha512-1hHV/Z4OEfMwpLO8rp7CvlhBDnjsC3CttJXIhBi+5Aj5r+MBvy4egg7wCbe//hSsT+RvDAG7s81tAvpL2XAE4w==",
+      "cpu": [
+        "ppc64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-riscv64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-riscv64/-/linux-riscv64-0.21.5.tgz",
+      "integrity": "sha512-2HdXDMd9GMgTGrPWnJzP2ALSokE/0O5HhTUvWIbD3YdjME8JwvSCnNGBnTThKGEB91OZhzrJ4qIIxk/SBmyDDA==",
+      "cpu": [
+        "riscv64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/linux-s390x": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/linux-s390x/-/linux-s390x-0.21.5.tgz",
+      "integrity": "sha512-zus5sxzqBJD3eXxwvjN1yQkRepANgxE9lgOW2qLnmr8ikMTphkjgXu1HR01K4FJg8h1kEEDAqDcZQtbrRnB41A==",
+      "cpu": [
+        "s390x"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
     "node_modules/@esbuild/linux-x64": {
       "version": "0.21.5",
       "cpu": [
@@ -333,6 +606,108 @@
         "node": ">=12"
       }
     },
+    "node_modules/@esbuild/netbsd-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/netbsd-x64/-/netbsd-x64-0.21.5.tgz",
+      "integrity": "sha512-Woi2MXzXjMULccIwMnLciyZH4nCIMpWQAs049KEeMvOcNADVxo0UBIQPfSmxB3CWKedngg7sWZdLvLczpe0tLg==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "netbsd"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/openbsd-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/openbsd-x64/-/openbsd-x64-0.21.5.tgz",
+      "integrity": "sha512-HLNNw99xsvx12lFBUwoT8EVCsSvRNDVxNpjZ7bPn947b8gJPzeHWyNVhFsaerc0n3TsbOINvRP2byTZ5LKezow==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "openbsd"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/sunos-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/sunos-x64/-/sunos-x64-0.21.5.tgz",
+      "integrity": "sha512-6+gjmFpfy0BHU5Tpptkuh8+uw3mnrvgs+dSPQXQOv3ekbordwnzTVEb4qnIvQcYXq6gzkyTnoZ9dZG+D4garKg==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "sunos"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/win32-arm64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-arm64/-/win32-arm64-0.21.5.tgz",
+      "integrity": "sha512-Z0gOTd75VvXqyq7nsl93zwahcTROgqvuAcYDUr+vOv8uHhNSKROyU961kgtCD1e95IqPKSQKH7tBTslnS3tA8A==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/win32-ia32": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-ia32/-/win32-ia32-0.21.5.tgz",
+      "integrity": "sha512-SWXFF1CL2RVNMaVs+BBClwtfZSvDgtL//G/smwAc5oVK/UPu2Gu9tIaRgFmYFFKrmg3SyAjSrElf0TiJ1v8fYA==",
+      "cpu": [
+        "ia32"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
+    "node_modules/@esbuild/win32-x64": {
+      "version": "0.21.5",
+      "resolved": "https://registry.npmjs.org/@esbuild/win32-x64/-/win32-x64-0.21.5.tgz",
+      "integrity": "sha512-tQd/1efJuzPC6rCFwEvLtci/xNFcTZknmXs98FYDfGE4wP9ClFV98nyKrzJKVPMhdDnjzLhdUyMX4PsQAPjwIw==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ],
+      "engines": {
+        "node": ">=12"
+      }
+    },
     "node_modules/@floating-ui/core": {
       "version": "1.7.5",
       "license": "MIT",
@@ -1280,6 +1655,277 @@
       "dev": true,
       "license": "MIT"
     },
+    "node_modules/@rollup/rollup-android-arm-eabi": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-android-arm-eabi/-/rollup-android-arm-eabi-4.60.1.tgz",
+      "integrity": "sha512-d6FinEBLdIiK+1uACUttJKfgZREXrF0Qc2SmLII7W2AD8FfiZ9Wjd+rD/iRuf5s5dWrr1GgwXCvPqOuDquOowA==",
+      "cpu": [
+        "arm"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ]
+    },
+    "node_modules/@rollup/rollup-android-arm64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-android-arm64/-/rollup-android-arm64-4.60.1.tgz",
+      "integrity": "sha512-YjG/EwIDvvYI1YvYbHvDz/BYHtkY4ygUIXHnTdLhG+hKIQFBiosfWiACWortsKPKU/+dUwQQCKQM3qrDe8c9BA==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "android"
+      ]
+    },
+    "node_modules/@rollup/rollup-darwin-arm64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-darwin-arm64/-/rollup-darwin-arm64-4.60.1.tgz",
+      "integrity": "sha512-mjCpF7GmkRtSJwon+Rq1N8+pI+8l7w5g9Z3vWj4T7abguC4Czwi3Yu/pFaLvA3TTeMVjnu3ctigusqWUfjZzvw==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ]
+    },
+    "node_modules/@rollup/rollup-darwin-x64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-darwin-x64/-/rollup-darwin-x64-4.60.1.tgz",
+      "integrity": "sha512-haZ7hJ1JT4e9hqkoT9R/19XW2QKqjfJVv+i5AGg57S+nLk9lQnJ1F/eZloRO3o9Scy9CM3wQ9l+dkXtcBgN5Ew==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ]
+    },
+    "node_modules/@rollup/rollup-freebsd-arm64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-freebsd-arm64/-/rollup-freebsd-arm64-4.60.1.tgz",
+      "integrity": "sha512-czw90wpQq3ZsAVBlinZjAYTKduOjTywlG7fEeWKUA7oCmpA8xdTkxZZlwNJKWqILlq0wehoZcJYfBvOyhPTQ6w==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ]
+    },
+    "node_modules/@rollup/rollup-freebsd-x64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-freebsd-x64/-/rollup-freebsd-x64-4.60.1.tgz",
+      "integrity": "sha512-KVB2rqsxTHuBtfOeySEyzEOB7ltlB/ux38iu2rBQzkjbwRVlkhAGIEDiiYnO2kFOkJp+Z7pUXKyrRRFuFUKt+g==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "freebsd"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-arm-gnueabihf": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-arm-gnueabihf/-/rollup-linux-arm-gnueabihf-4.60.1.tgz",
+      "integrity": "sha512-L+34Qqil+v5uC0zEubW7uByo78WOCIrBvci69E7sFASRl0X7b/MB6Cqd1lky/CtcSVTydWa2WZwFuWexjS5o6g==",
+      "cpu": [
+        "arm"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-arm-musleabihf": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-arm-musleabihf/-/rollup-linux-arm-musleabihf-4.60.1.tgz",
+      "integrity": "sha512-n83O8rt4v34hgFzlkb1ycniJh7IR5RCIqt6mz1VRJD6pmhRi0CXdmfnLu9dIUS6buzh60IvACM842Ffb3xd6Gg==",
+      "cpu": [
+        "arm"
+      ],
+      "dev": true,
+      "libc": [
+        "musl"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-arm64-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-arm64-gnu/-/rollup-linux-arm64-gnu-4.60.1.tgz",
+      "integrity": "sha512-Nql7sTeAzhTAja3QXeAI48+/+GjBJ+QmAH13snn0AJSNL50JsDqotyudHyMbO2RbJkskbMbFJfIJKWA6R1LCJQ==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-arm64-musl": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-arm64-musl/-/rollup-linux-arm64-musl-4.60.1.tgz",
+      "integrity": "sha512-+pUymDhd0ys9GcKZPPWlFiZ67sTWV5UU6zOJat02M1+PiuSGDziyRuI/pPue3hoUwm2uGfxdL+trT6Z9rxnlMA==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "libc": [
+        "musl"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-loong64-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-loong64-gnu/-/rollup-linux-loong64-gnu-4.60.1.tgz",
+      "integrity": "sha512-VSvgvQeIcsEvY4bKDHEDWcpW4Yw7BtlKG1GUT4FzBUlEKQK0rWHYBqQt6Fm2taXS+1bXvJT6kICu5ZwqKCnvlQ==",
+      "cpu": [
+        "loong64"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-loong64-musl": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-loong64-musl/-/rollup-linux-loong64-musl-4.60.1.tgz",
+      "integrity": "sha512-4LqhUomJqwe641gsPp6xLfhqWMbQV04KtPp7/dIp0nzPxAkNY1AbwL5W0MQpcalLYk07vaW9Kp1PBhdpZYYcEw==",
+      "cpu": [
+        "loong64"
+      ],
+      "dev": true,
+      "libc": [
+        "musl"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-ppc64-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-ppc64-gnu/-/rollup-linux-ppc64-gnu-4.60.1.tgz",
+      "integrity": "sha512-tLQQ9aPvkBxOc/EUT6j3pyeMD6Hb8QF2BTBnCQWP/uu1lhc9AIrIjKnLYMEroIz/JvtGYgI9dF3AxHZNaEH0rw==",
+      "cpu": [
+        "ppc64"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-ppc64-musl": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-ppc64-musl/-/rollup-linux-ppc64-musl-4.60.1.tgz",
+      "integrity": "sha512-RMxFhJwc9fSXP6PqmAz4cbv3kAyvD1etJFjTx4ONqFP9DkTkXsAMU4v3Vyc5BgzC+anz7nS/9tp4obsKfqkDHg==",
+      "cpu": [
+        "ppc64"
+      ],
+      "dev": true,
+      "libc": [
+        "musl"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-riscv64-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-riscv64-gnu/-/rollup-linux-riscv64-gnu-4.60.1.tgz",
+      "integrity": "sha512-QKgFl+Yc1eEk6MmOBfRHYF6lTxiiiV3/z/BRrbSiW2I7AFTXoBFvdMEyglohPj//2mZS4hDOqeB0H1ACh3sBbg==",
+      "cpu": [
+        "riscv64"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-riscv64-musl": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-riscv64-musl/-/rollup-linux-riscv64-musl-4.60.1.tgz",
+      "integrity": "sha512-RAjXjP/8c6ZtzatZcA1RaQr6O1TRhzC+adn8YZDnChliZHviqIjmvFwHcxi4JKPSDAt6Uhf/7vqcBzQJy0PDJg==",
+      "cpu": [
+        "riscv64"
+      ],
+      "dev": true,
+      "libc": [
+        "musl"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
+    "node_modules/@rollup/rollup-linux-s390x-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-linux-s390x-gnu/-/rollup-linux-s390x-gnu-4.60.1.tgz",
+      "integrity": "sha512-wcuocpaOlaL1COBYiA89O6yfjlp3RwKDeTIA0hM7OpmhR1Bjo9j31G1uQVpDlTvwxGn2nQs65fBFL5UFd76FcQ==",
+      "cpu": [
+        "s390x"
+      ],
+      "dev": true,
+      "libc": [
+        "glibc"
+      ],
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "linux"
+      ]
+    },
     "node_modules/@rollup/rollup-linux-x64-gnu": {
       "version": "4.60.1",
       "cpu": [
@@ -1304,6 +1950,90 @@
         "linux"
       ]
     },
+    "node_modules/@rollup/rollup-openbsd-x64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-openbsd-x64/-/rollup-openbsd-x64-4.60.1.tgz",
+      "integrity": "sha512-cl0w09WsCi17mcmWqqglez9Gk8isgeWvoUZ3WiJFYSR3zjBQc2J5/ihSjpl+VLjPqjQ/1hJRcqBfLjssREQILw==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "openbsd"
+      ]
+    },
+    "node_modules/@rollup/rollup-openharmony-arm64": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-openharmony-arm64/-/rollup-openharmony-arm64-4.60.1.tgz",
+      "integrity": "sha512-4Cv23ZrONRbNtbZa37mLSueXUCtN7MXccChtKpUnQNgF010rjrjfHx3QxkS2PI7LqGT5xXyYs1a7LbzAwT0iCA==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "openharmony"
+      ]
+    },
+    "node_modules/@rollup/rollup-win32-arm64-msvc": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-win32-arm64-msvc/-/rollup-win32-arm64-msvc-4.60.1.tgz",
+      "integrity": "sha512-i1okWYkA4FJICtr7KpYzFpRTHgy5jdDbZiWfvny21iIKky5YExiDXP+zbXzm3dUcFpkEeYNHgQ5fuG236JPq0g==",
+      "cpu": [
+        "arm64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ]
+    },
+    "node_modules/@rollup/rollup-win32-ia32-msvc": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-win32-ia32-msvc/-/rollup-win32-ia32-msvc-4.60.1.tgz",
+      "integrity": "sha512-u09m3CuwLzShA0EYKMNiFgcjjzwqtUMLmuCJLeZWjjOYA3IT2Di09KaxGBTP9xVztWyIWjVdsB2E9goMjZvTQg==",
+      "cpu": [
+        "ia32"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ]
+    },
+    "node_modules/@rollup/rollup-win32-x64-gnu": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-win32-x64-gnu/-/rollup-win32-x64-gnu-4.60.1.tgz",
+      "integrity": "sha512-k+600V9Zl1CM7eZxJgMyTUzmrmhB/0XZnF4pRypKAlAgxmedUA+1v9R+XOFv56W4SlHEzfeMtzujLJD22Uz5zg==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ]
+    },
+    "node_modules/@rollup/rollup-win32-x64-msvc": {
+      "version": "4.60.1",
+      "resolved": "https://registry.npmjs.org/@rollup/rollup-win32-x64-msvc/-/rollup-win32-x64-msvc-4.60.1.tgz",
+      "integrity": "sha512-lWMnixq/QzxyhTV6NjQJ4SFo1J6PvOX8vUx5Wb4bBPsEb+8xZ89Bz6kOXpfXj9ak9AHTQVQzlgzBEc1SyM27xQ==",
+      "cpu": [
+        "x64"
+      ],
+      "dev": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "win32"
+      ]
+    },
     "node_modules/@tailwindcss/typography": {
       "version": "0.5.19",
       "dev": true,
@@ -2309,6 +3039,21 @@
         "url": "https://github.com/sponsors/rawify"
       }
     },
+    "node_modules/fsevents": {
+      "version": "2.3.3",
+      "resolved": "https://registry.npmjs.org/fsevents/-/fsevents-2.3.3.tgz",
+      "integrity": "sha512-5xoDfX+fL7faATnagmWPpbFtwh/R77WmMMqqHGS65C3vvB0YHrgF+B1YmZ3441tMj5n63k0212XNoJwzlhffQw==",
+      "dev": true,
+      "hasInstallScript": true,
+      "license": "MIT",
+      "optional": true,
+      "os": [
+        "darwin"
+      ],
+      "engines": {
+        "node": "^8.16.0 || ^10.6.0 || >=11.0.0"
+      }
+    },
     "node_modules/function-bind": {
       "version": "1.1.2",
       "dev": true,
@@ -3178,6 +3923,20 @@
         "url": "https://opencollective.com/unified"
       }
     },
+    "node_modules/mdast-util-newline-to-break": {
+      "version": "2.0.0",
+      "resolved": "https://registry.npmjs.org/mdast-util-newline-to-break/-/mdast-util-newline-to-break-2.0.0.tgz",
+      "integrity": "sha512-MbgeFca0hLYIEx/2zGsszCSEJJ1JSCdiY5xQxRcLDDGa8EPvlLPupJ4DSajbMPAnC0je8jfb9TiUATnxxrHUog==",
+      "license": "MIT",
+      "dependencies": {
+        "@types/mdast": "^4.0.0",
+        "mdast-util-find-and-replace": "^3.0.0"
+      },
+      "funding": {
+        "type": "opencollective",
+        "url": "https://opencollective.com/unified"
+      }
+    },
     "node_modules/mdast-util-phrasing": {
       "version": "4.1.0",
       "license": "MIT",
@@ -4397,6 +5156,21 @@
         "url": "https://opencollective.com/unified"
       }
     },
+    "node_modules/remark-breaks": {
+      "version": "4.0.0",
+      "resolved": "https://registry.npmjs.org/remark-breaks/-/remark-breaks-4.0.0.tgz",
+      "integrity": "sha512-IjEjJOkH4FuJvHZVIW0QCDWxcG96kCq7An/KVH2NfJe6rKZU2AsHeB3OEjPNRxi4QC34Xdx7I2KGYn6IpT7gxQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@types/mdast": "^4.0.0",
+        "mdast-util-newline-to-break": "^2.0.0",
+        "unified": "^11.0.0"
+      },
+      "funding": {
+        "type": "opencollective",
+        "url": "https://opencollective.com/unified"
+      }
+    },
     "node_modules/remark-gfm": {
       "version": "4.0.1",
       "license": "MIT",

webui/package.json

@@ -30,6 +30,7 @@
     "react-markdown": "^9.0.1",
     "react-syntax-highlighter": "^15.6.1",
     "rehype-katex": "^7.0.1",
+    "remark-breaks": "^4.0.0",
     "remark-gfm": "^4.0.0",
     "remark-math": "^6.0.0",
     "tailwind-merge": "^2.6.0"

webui/src/App.tsx

@@ -1,13 +1,17 @@
 import { useCallback, useEffect, useMemo, useRef, useState } from "react";
 import { useTranslation } from "react-i18next";
 import { DeleteConfirm } from "@/components/DeleteConfirm";
+import { RenameChatDialog } from "@/components/RenameChatDialog";
 import { Sidebar } from "@/components/Sidebar";
+import { SessionSearchDialog } from "@/components/SessionSearchDialog";
 import { SettingsView } from "@/components/settings/SettingsView";
 import { ThreadShell } from "@/components/thread/ThreadShell";
-import { Sheet, SheetContent } from "@/components/ui/sheet";
+import { Sheet, SheetContent, SheetTitle } from "@/components/ui/sheet";
 
 import { useSessions } from "@/hooks/useSessions";
-import { useTheme } from "@/hooks/useTheme";
+import { useDeferredTitleRefresh } from "@/hooks/useDeferredTitleRefresh";
+import { useSidebarState } from "@/hooks/useSidebarState";
+import { ThemeProvider, useTheme } from "@/hooks/useTheme";
 import { cn } from "@/lib/utils";
 import {
   clearSavedSecret,
@@ -16,6 +20,7 @@ import {
   loadSavedSecret,
   saveSecret,
 } from "@/lib/bootstrap";
+import { deriveTitle } from "@/lib/format";
 import { NanobotClient } from "@/lib/nanobot-client";
 import { ClientProvider, useClient } from "@/providers/ClientProvider";
 import type { ChatSummary } from "@/lib/types";
@@ -30,14 +35,31 @@ type BootState =
       status: "ready";
       client: NanobotClient;
       token: string;
+      tokenExpiresAt: number;
       modelName: string | null;
     };
 
 const SIDEBAR_STORAGE_KEY = "nanobot-webui.sidebar";
+const COMPLETED_RUNS_STORAGE_KEY = "nanobot-webui.sidebar.completed-runs.v1";
 const RESTART_STARTED_KEY = "nanobot-webui.restartStartedAt";
 const SIDEBAR_WIDTH = 272;
+const TOKEN_REFRESH_MARGIN_MS = 30_000;
+const TOKEN_REFRESH_MIN_DELAY_MS = 5_000;
 type ShellView = "chat" | "settings";
 
+function bootstrapTokenExpiresAt(expiresInSeconds: number): number {
+  return Date.now() + Math.max(0, expiresInSeconds) * 1000;
+}
+
+function tokenRefreshDelayMs(expiresAt: number): number {
+  const remaining = Math.max(0, expiresAt - Date.now());
+  const margin = Math.min(
+    TOKEN_REFRESH_MARGIN_MS,
+    Math.max(1_000, remaining / 2),
+  );
+  return Math.max(TOKEN_REFRESH_MIN_DELAY_MS, remaining - margin);
+}
+
 function AuthForm({
   failed,
   onSecret,
@@ -103,9 +125,33 @@ function readSidebarOpen(): boolean {
   }
 }
 
+function readCompletedRunChatIds(): Set<string> {
+  if (typeof window === "undefined") return new Set();
+  try {
+    const raw = window.localStorage.getItem(COMPLETED_RUNS_STORAGE_KEY);
+    const parsed = raw ? JSON.parse(raw) : [];
+    if (!Array.isArray(parsed)) return new Set();
+    return new Set(parsed.filter((item): item is string => typeof item === "string"));
+  } catch {
+    return new Set();
+  }
+}
+
+function writeCompletedRunChatIds(chatIds: Set<string>): void {
+  try {
+    window.localStorage.setItem(
+      COMPLETED_RUNS_STORAGE_KEY,
+      JSON.stringify(Array.from(chatIds)),
+    );
+  } catch {
+    // ignore storage errors (private mode, etc.)
+  }
+}
+
 export default function App() {
   const { t } = useTranslation();
   const [state, setState] = useState<BootState>({ status: "loading" });
+  const bootstrapSecretRef = useRef("");
 
   const bootstrapWithSecret = useCallback(
     (secret: string) => {
@@ -117,22 +163,37 @@ export default function App() {
           if (cancelled) return;
           if (secret) saveSecret(secret);
           const url = deriveWsUrl(boot.ws_path, boot.token);
-          const client = new NanobotClient({
+          let client: NanobotClient;
+          client = new NanobotClient({
             url,
             onReauth: async () => {
               try {
-                const refreshed = await fetchBootstrap("", secret);
-                return deriveWsUrl(refreshed.ws_path, refreshed.token);
+                const refreshed = await fetchBootstrap("", bootstrapSecretRef.current);
+                const refreshedUrl = deriveWsUrl(refreshed.ws_path, refreshed.token);
+                const tokenExpiresAt = bootstrapTokenExpiresAt(refreshed.expires_in);
+                setState((current) =>
+                  current.status === "ready" && current.client === client
+                    ? {
+                        ...current,
+                        token: refreshed.token,
+                        tokenExpiresAt,
+                        modelName: refreshed.model_name ?? current.modelName,
+                      }
+                    : current,
+                );
+                return refreshedUrl;
               } catch {
                 return null;
               }
             },
           });
+          bootstrapSecretRef.current = secret;
           client.connect();
           setState({
             status: "ready",
             client,
             token: boot.token,
+            tokenExpiresAt: bootstrapTokenExpiresAt(boot.expires_in),
             modelName: boot.model_name ?? null,
           });
         } catch (e) {
@@ -152,6 +213,35 @@ export default function App() {
     [],
   );
 
+  useEffect(() => {
+    if (state.status !== "ready") return;
+    const client = state.client;
+    const timer = window.setTimeout(async () => {
+      try {
+        const boot = await fetchBootstrap("", bootstrapSecretRef.current);
+        const url = deriveWsUrl(boot.ws_path, boot.token);
+        const tokenExpiresAt = bootstrapTokenExpiresAt(boot.expires_in);
+        client.updateUrl(url);
+        setState((current) =>
+          current.status === "ready" && current.client === client
+            ? {
+                ...current,
+                token: boot.token,
+                tokenExpiresAt,
+                modelName: boot.model_name ?? current.modelName,
+              }
+            : current,
+        );
+      } catch (e) {
+        const msg = (e as Error).message;
+        if (msg.includes("HTTP 401") || msg.includes("HTTP 403")) {
+          setState({ status: "auth", failed: true });
+        }
+      }
+    }, tokenRefreshDelayMs(state.tokenExpiresAt));
+    return () => window.clearTimeout(timer);
+  }, [state]);
+
   useEffect(() => {
     const saved = loadSavedSecret();
     return bootstrapWithSecret(saved);
@@ -219,23 +309,39 @@ export default function App() {
   );
 }
 
-function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName: string | null) => void; onLogout: () => void }) {
+function Shell({
+  onModelNameChange,
+  onLogout,
+}: {
+  onModelNameChange: (modelName: string | null) => void;
+  onLogout: () => void;
+}) {
   const { t, i18n } = useTranslation();
   const { client } = useClient();
   const { theme, toggle } = useTheme();
   const { sessions, loading, refresh, createChat, deleteChat } = useSessions();
+  const { state: sidebarState, update: updateSidebarState } =
+    useSidebarState(sessions, !loading);
   const [activeKey, setActiveKey] = useState<string | null>(null);
   const [view, setView] = useState<ShellView>("chat");
   const [desktopSidebarOpen, setDesktopSidebarOpen] =
     useState<boolean>(readSidebarOpen);
   const [mobileSidebarOpen, setMobileSidebarOpen] = useState(false);
+  const [sessionSearchOpen, setSessionSearchOpen] = useState(false);
   const [pendingDelete, setPendingDelete] = useState<{
     key: string;
     label: string;
   } | null>(null);
+  const [pendingRename, setPendingRename] = useState<{
+    key: string;
+    label: string;
+  } | null>(null);
   const restartSawDisconnectRef = useRef(false);
   const [restartToast, setRestartToast] = useState<string | null>(null);
   const [isRestarting, setIsRestarting] = useState(false);
+  const [runningChatIds, setRunningChatIds] = useState<Set<string>>(() => new Set());
+  const [completedChatIds, setCompletedChatIds] = useState<Set<string>>(readCompletedRunChatIds);
+  const runningChatIdsRef = useRef<Set<string>>(new Set());
 
   useEffect(() => {
     try {
@@ -248,12 +354,58 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
     }
   }, [desktopSidebarOpen]);
 
-
+  useEffect(() => {
+    writeCompletedRunChatIds(completedChatIds);
+  }, [completedChatIds]);
 
   const activeSession = useMemo<ChatSummary | null>(() => {
     if (!activeKey) return null;
     return sessions.find((s) => s.key === activeKey) ?? null;
   }, [sessions, activeKey]);
+  const runningChatIdList = useMemo(() => Array.from(runningChatIds), [runningChatIds]);
+  const completedChatIdList = useMemo(() => Array.from(completedChatIds), [completedChatIds]);
+
+  useEffect(() => {
+    if (loading) return;
+    const knownChatIds = new Set(sessions.map((session) => session.chatId));
+    setCompletedChatIds((current) => {
+      const next = new Set(
+        Array.from(current).filter((chatId) => knownChatIds.has(chatId)),
+      );
+      return next.size === current.size ? current : next;
+    });
+  }, [loading, sessions]);
+
+  useEffect(() => {
+    if (loading) return;
+    const activeRunIds = sessions
+      .filter((session) => typeof session.runStartedAt === "number")
+      .map((session) => session.chatId);
+    if (activeRunIds.length === 0) return;
+
+    for (const chatId of activeRunIds) {
+      client.attach(chatId);
+    }
+    setRunningChatIds((current) => {
+      let changed = false;
+      const next = new Set(current);
+      for (const chatId of activeRunIds) {
+        if (!next.has(chatId)) changed = true;
+        next.add(chatId);
+      }
+      if (!changed) return current;
+      runningChatIdsRef.current = next;
+      return next;
+    });
+    setCompletedChatIds((current) => {
+      let changed = false;
+      const next = new Set(current);
+      for (const chatId of activeRunIds) {
+        if (next.delete(chatId)) changed = true;
+      }
+      return changed ? next : current;
+    });
+  }, [client, loading, sessions]);
 
   const closeDesktopSidebar = useCallback(() => {
     setDesktopSidebarOpen(false);
@@ -295,14 +447,129 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
 
   const onSelectChat = useCallback(
     (key: string) => {
+      const selectedChatId = sessions.find((session) => session.key === key)?.chatId;
+      if (selectedChatId) {
+        setCompletedChatIds((current) => {
+          if (!current.has(selectedChatId)) return current;
+          const next = new Set(current);
+          next.delete(selectedChatId);
+          return next;
+        });
+      }
       setActiveKey(key);
       setView("chat");
       setMobileSidebarOpen(false);
     },
-    [],
+    [sessions],
+  );
+
+  const onTogglePin = useCallback(
+    (key: string) => {
+      void updateSidebarState((current) => {
+        const pinned = new Set(current.pinned_keys);
+        if (pinned.has(key)) {
+          pinned.delete(key);
+        } else {
+          pinned.add(key);
+        }
+        return {
+          ...current,
+          pinned_keys: Array.from(pinned),
+        };
+      });
+    },
+    [updateSidebarState],
+  );
+
+  const onRequestRename = useCallback((key: string, label: string) => {
+    setPendingRename({ key, label });
+  }, []);
+
+  const onConfirmRename = useCallback(
+    (title: string) => {
+      if (!pendingRename) return;
+      const key = pendingRename.key;
+      setPendingRename(null);
+      void updateSidebarState((current) => {
+        const titleOverrides = { ...current.title_overrides };
+        const cleaned = title.trim();
+        if (cleaned) {
+          titleOverrides[key] = cleaned;
+        } else {
+          delete titleOverrides[key];
+        }
+        return {
+          ...current,
+          title_overrides: titleOverrides,
+        };
+      });
+    },
+    [pendingRename, updateSidebarState],
+  );
+
+  const onToggleArchive = useCallback(
+    (key: string) => {
+      void updateSidebarState((current) => {
+        const archived = new Set(current.archived_keys);
+        const pinned = current.pinned_keys.filter((item) => item !== key);
+        if (archived.has(key)) {
+          archived.delete(key);
+        } else {
+          archived.add(key);
+        }
+        return {
+          ...current,
+          pinned_keys: pinned,
+          archived_keys: Array.from(archived),
+        };
+      });
+      if (activeKey === key && !sidebarState.archived_keys.includes(key)) {
+        const archived = new Set([...sidebarState.archived_keys, key]);
+        const next = sessions.find((session) => !archived.has(session.key));
+        setActiveKey(next?.key ?? null);
+      }
+    },
+    [activeKey, sessions, sidebarState.archived_keys, updateSidebarState],
+  );
+
+  const onToggleArchived = useCallback(() => {
+    void updateSidebarState((current) => ({
+      ...current,
+      view: {
+        ...current.view,
+        show_archived: !current.view.show_archived,
+      },
+    }));
+  }, [updateSidebarState]);
+
+  const onUpdateSidebarView = useCallback(
+    (viewUpdate: Partial<typeof sidebarState.view>) => {
+      void updateSidebarState((current) => ({
+        ...current,
+        view: {
+          ...current.view,
+          ...viewUpdate,
+        },
+      }));
+    },
+    [updateSidebarState],
+  );
+
+  const onOpenSessionSearch = useCallback(() => {
+    setMobileSidebarOpen(false);
+    setSessionSearchOpen(true);
+  }, []);
+
+  const onSelectSearchResult = useCallback(
+    (key: string) => {
+      setSessionSearchOpen(false);
+      onSelectChat(key);
+    },
+    [onSelectChat],
   );
 
   const onOpenSettings = useCallback(() => {
+    setSessionSearchOpen(false);
     setView("settings");
     setMobileSidebarOpen(false);
   }, []);
@@ -336,6 +603,35 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
     });
   }, [client, onModelNameChange]);
 
+  useEffect(() => {
+    return client.onRunStatus((chatId, startedAt) => {
+      if (startedAt != null) {
+        const nextRunning = new Set(runningChatIdsRef.current);
+        nextRunning.add(chatId);
+        runningChatIdsRef.current = nextRunning;
+        setRunningChatIds(nextRunning);
+        setCompletedChatIds((current) => {
+          if (!current.has(chatId)) return current;
+          const next = new Set(current);
+          next.delete(chatId);
+          return next;
+        });
+        return;
+      }
+
+      if (!runningChatIdsRef.current.has(chatId)) return;
+      const nextRunning = new Set(runningChatIdsRef.current);
+      nextRunning.delete(chatId);
+      runningChatIdsRef.current = nextRunning;
+      setRunningChatIds(nextRunning);
+      setCompletedChatIds((current) => {
+        const next = new Set(current);
+        next.add(chatId);
+        return next;
+      });
+    });
+  }, [client]);
+
   useEffect(() => {
     return client.onStatus((status) => {
       let startedAt = 0;
@@ -362,9 +658,7 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
     });
   }, [client, t]);
 
-  const onTurnEnd = useCallback(() => {
-    void refresh();
-  }, [refresh]);
+  const onTurnEnd = useDeferredTitleRefresh(activeSession, refresh);
 
   const onConfirmDelete = useCallback(async () => {
     if (!pendingDelete) return;
@@ -385,9 +679,9 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
   }, [pendingDelete, deleteChat, activeKey, sessions]);
 
   const headerTitle = activeSession
-    ? activeSession.title ||
-      activeSession.preview ||
-      t("chat.fallbackTitle", { id: activeSession.chatId.slice(0, 6) })
+    ? sidebarState.title_overrides[activeSession.key] ||
+      activeSession.title ||
+      deriveTitle(activeSession.preview, t("chat.newChat"))
     : t("app.brand");
 
   useEffect(() => {
@@ -410,98 +704,138 @@ function Shell({ onModelNameChange, onLogout }: { onModelNameChange: (modelName:
     onSelect: onSelectChat,
     onRequestDelete: (key: string, label: string) =>
       setPendingDelete({ key, label }),
+    onTogglePin,
+    onRequestRename,
+    onToggleArchive,
     onOpenSettings,
+    onOpenSearch: onOpenSessionSearch,
+    onToggleArchived,
+    onUpdateView: onUpdateSidebarView,
+    pinnedKeys: sidebarState.pinned_keys,
+    archivedKeys: sidebarState.archived_keys,
+    titleOverrides: sidebarState.title_overrides,
+    runningChatIds: runningChatIdList,
+    completedChatIds: completedChatIdList,
+    viewState: sidebarState.view,
+    showArchived: sidebarState.view.show_archived,
+    archivedCount: sidebarState.archived_keys.length,
   };
   const showMainSidebar = view !== "settings";
 
   return (
-    <div className="relative flex h-full w-full overflow-hidden">
-      {/* Desktop sidebar: in normal flow, so the thread area width stays honest. */}
-      {showMainSidebar ? (
-        <aside
-          className={cn(
-            "relative z-20 hidden shrink-0 overflow-hidden lg:block",
-            "transition-[width] duration-300 ease-out",
-          )}
-          style={{ width: desktopSidebarOpen ? SIDEBAR_WIDTH : 0 }}
-        >
+    <ThemeProvider theme={theme}>
+      <div className="relative flex h-full w-full overflow-hidden">
+        {/* Desktop sidebar: in normal flow, so the thread area width stays honest. */}
+        {showMainSidebar ? (
+          <aside
+            className={cn(
+              "relative z-20 hidden shrink-0 overflow-hidden lg:block",
+              "transition-[width] duration-300 ease-out",
+            )}
+            style={{ width: desktopSidebarOpen ? SIDEBAR_WIDTH : 0 }}
+          >
+            <div
+              className={cn(
+                "absolute inset-y-0 left-0 h-full overflow-hidden bg-sidebar shadow-inner-right",
+                "transition-transform duration-300 ease-out",
+                desktopSidebarOpen ? "translate-x-0" : "-translate-x-full",
+              )}
+              style={{ width: SIDEBAR_WIDTH }}
+            >
+              <Sidebar {...sidebarProps} onCollapse={closeDesktopSidebar} />
+            </div>
+          </aside>
+        ) : null}
+
+        {showMainSidebar ? (
+          <Sheet
+            open={mobileSidebarOpen}
+            onOpenChange={(open) => setMobileSidebarOpen(open)}
+          >
+            <SheetContent
+              side="left"
+              showCloseButton={false}
+              aria-describedby={undefined}
+              className="p-0 lg:hidden"
+              style={{ width: SIDEBAR_WIDTH, maxWidth: SIDEBAR_WIDTH }}
+            >
+              <SheetTitle className="sr-only">{t("sidebar.navigation")}</SheetTitle>
+              <Sidebar
+                {...sidebarProps}
+                onCollapse={closeMobileSidebar}
+                containActionMenus
+              />
+            </SheetContent>
+          </Sheet>
+        ) : null}
+
+        {showMainSidebar ? (
+          <SessionSearchDialog
+            open={sessionSearchOpen}
+            onOpenChange={setSessionSearchOpen}
+            sessions={sessions}
+            activeKey={activeKey}
+            loading={loading}
+            titleOverrides={sidebarState.title_overrides}
+            onSelect={onSelectSearchResult}
+          />
+        ) : null}
+
+        <main className="relative flex h-full min-w-0 flex-1 flex-col">
           <div
             className={cn(
-              "absolute inset-y-0 left-0 h-full overflow-hidden bg-sidebar shadow-inner-right",
-              "transition-transform duration-300 ease-out",
-              desktopSidebarOpen ? "translate-x-0" : "-translate-x-full",
+              "absolute inset-0 flex flex-col",
+              view === "settings" && "invisible pointer-events-none",
             )}
-            style={{ width: SIDEBAR_WIDTH }}
           >
-            <Sidebar {...sidebarProps} onCollapse={closeDesktopSidebar} />
-          </div>
-        </aside>
-      ) : null}
-
-      {showMainSidebar ? (
-        <Sheet
-          open={mobileSidebarOpen}
-          onOpenChange={(open) => setMobileSidebarOpen(open)}
-        >
-          <SheetContent
-            side="left"
-            showCloseButton={false}
-            className="p-0 lg:hidden"
-            style={{ width: SIDEBAR_WIDTH, maxWidth: SIDEBAR_WIDTH }}
-          >
-            <Sidebar {...sidebarProps} onCollapse={closeMobileSidebar} />
-          </SheetContent>
-        </Sheet>
-      ) : null}
-
-      <main className="relative flex h-full min-w-0 flex-1 flex-col">
-        <div
-          className={cn(
-            "absolute inset-0 flex flex-col",
-            view === "settings" && "invisible pointer-events-none",
-          )}
-        >
-          <ThreadShell
-            session={activeSession}
-            title={headerTitle}
-            onToggleSidebar={toggleSidebar}
-            onNewChat={onNewChat}
-            onCreateChat={onCreateChat}
-            onTurnEnd={onTurnEnd}
-            theme={theme}
-            onToggleTheme={toggle}
-            hideSidebarToggleOnDesktop={desktopSidebarOpen}
-          />
-        </div>
-        {view === "settings" && (
-          <div className="absolute inset-0 flex flex-col">
-            <SettingsView
+            <ThreadShell
+              session={activeSession}
+              title={headerTitle}
+              onToggleSidebar={toggleSidebar}
+              onNewChat={onNewChat}
+              onCreateChat={onCreateChat}
+              onTurnEnd={onTurnEnd}
               theme={theme}
               onToggleTheme={toggle}
-              onBackToChat={onBackToChat}
-              onModelNameChange={onModelNameChange}
-              onLogout={onLogout}
-              onRestart={onRestart}
-              isRestarting={isRestarting}
+              hideSidebarToggleOnDesktop={desktopSidebarOpen}
             />
           </div>
-        )}
-      </main>
+          {view === "settings" && (
+            <div className="absolute inset-0 flex flex-col">
+              <SettingsView
+                theme={theme}
+                onToggleTheme={toggle}
+                onBackToChat={onBackToChat}
+                onModelNameChange={onModelNameChange}
+                onLogout={onLogout}
+                onRestart={onRestart}
+                isRestarting={isRestarting}
+              />
+            </div>
+          )}
+        </main>
 
-      <DeleteConfirm
-        open={!!pendingDelete}
-        title={pendingDelete?.label ?? ""}
-        onCancel={() => setPendingDelete(null)}
-        onConfirm={onConfirmDelete}
-      />
-      {restartToast ? (
-        <div
-          role="status"
-          className="fixed left-1/2 top-4 z-50 -translate-x-1/2 rounded-full border border-border/70 bg-popover px-4 py-2 text-sm font-medium text-popover-foreground shadow-lg"
-        >
-          {restartToast}
-        </div>
-      ) : null}
-    </div>
+        <DeleteConfirm
+          open={!!pendingDelete}
+          title={pendingDelete?.label ?? ""}
+          onCancel={() => setPendingDelete(null)}
+          onConfirm={onConfirmDelete}
+        />
+        <RenameChatDialog
+          open={!!pendingRename}
+          title={pendingRename?.label ?? ""}
+          onCancel={() => setPendingRename(null)}
+          onConfirm={onConfirmRename}
+        />
+        {restartToast ? (
+          <div
+            role="status"
+            className="fixed left-1/2 top-4 z-50 -translate-x-1/2 rounded-full border border-border/70 bg-popover px-4 py-2 text-sm font-medium text-popover-foreground shadow-lg"
+          >
+            {restartToast}
+          </div>
+        ) : null}
+      </div>
+    </ThemeProvider>
   );
 }

webui/src/components/ChatList.tsx

@@ -1,4 +1,12 @@
-import { MoreHorizontal, Trash2 } from "lucide-react";
+import {
+  Archive,
+  ArchiveRestore,
+  MoreHorizontal,
+  Pencil,
+  Pin,
+  PinOff,
+  Trash2,
+} from "lucide-react";
 import { useTranslation } from "react-i18next";
 
 import {
@@ -7,14 +15,29 @@ import {
   DropdownMenuItem,
   DropdownMenuTrigger,
 } from "@/components/ui/dropdown-menu";
+import { deriveTitle, relativeTime } from "@/lib/format";
 import { cn } from "@/lib/utils";
-import type { ChatSummary } from "@/lib/types";
+import type { ChatSummary, SidebarDensity, SidebarSortMode } from "@/lib/types";
 
 interface ChatListProps {
   sessions: ChatSummary[];
   activeKey: string | null;
   onSelect: (key: string) => void;
   onRequestDelete: (key: string, label: string) => void;
+  onTogglePin: (key: string) => void;
+  onRequestRename: (key: string, label: string) => void;
+  onToggleArchive: (key: string) => void;
+  pinnedKeys?: string[];
+  archivedKeys?: string[];
+  titleOverrides?: Record<string, string>;
+  runningChatIds?: string[];
+  completedChatIds?: string[];
+  density?: SidebarDensity;
+  showPreviews?: boolean;
+  showTimestamps?: boolean;
+  sort?: SidebarSortMode;
+  showArchived?: boolean;
+  actionMenuPortalContainer?: HTMLElement | null;
   loading?: boolean;
   emptyLabel?: string;
 }
@@ -24,6 +47,20 @@ export function ChatList({
   activeKey,
   onSelect,
   onRequestDelete,
+  onTogglePin,
+  onRequestRename,
+  onToggleArchive,
+  pinnedKeys = [],
+  archivedKeys = [],
+  titleOverrides = {},
+  runningChatIds = [],
+  completedChatIds = [],
+  density = "comfortable",
+  showPreviews = false,
+  showTimestamps = false,
+  sort = "updated_desc",
+  showArchived = false,
+  actionMenuPortalContainer,
   loading,
   emptyLabel,
 }: ChatListProps) {
@@ -45,10 +82,25 @@ export function ChatList({
   }
 
   const groups = groupSessions(sessions, {
+    pinned: t("chat.groups.pinned"),
+    all: t("chat.groups.all"),
     today: t("chat.groups.today"),
     yesterday: t("chat.groups.yesterday"),
     earlier: t("chat.groups.earlier"),
+    archived: t("chat.groups.archived"),
+    fallbackTitle: t("chat.newChat"),
+  }, {
+    pinnedKeys,
+    archivedKeys,
+    titleOverrides,
+    showArchived,
+    sort,
   });
+  const pinned = new Set(pinnedKeys);
+  const archived = new Set(archivedKeys);
+  const running = new Set(runningChatIds);
+  const completed = new Set(completedChatIds);
+  const compact = density === "compact";
 
   return (
     <div className="h-full min-h-0 min-w-0 overflow-x-hidden overflow-y-auto overscroll-contain">
@@ -64,13 +116,30 @@ export function ChatList({
                 const fallbackTitle = t("chat.fallbackTitle", {
                   id: s.chatId.slice(0, 6),
                 });
-                const rawLabel = (s.title || s.preview)?.trim();
-                const title = rawLabel || fallbackTitle;
+                const generatedTitle = s.title?.trim() || "";
+                const title = displayTitle(s, titleOverrides, t("chat.newChat"));
+                const tooltipTitle =
+                  titleOverrides[s.key]?.trim() ||
+                  generatedTitle ||
+                  deriveTitle(s.preview, fallbackTitle);
+                const isPinned = pinned.has(s.key);
+                const isArchived = archived.has(s.key);
+                const preview = s.preview.trim();
+                const showPreview = showPreviews && preview && preview !== title;
+                const timestamp = showTimestamps
+                  ? relativeTime(s.updatedAt ?? s.createdAt)
+                  : "";
+                const activityState = running.has(s.chatId)
+                  ? "running"
+                  : completed.has(s.chatId)
+                    ? "complete"
+                    : null;
                 return (
                   <li key={s.key} className="min-w-0">
                     <div
                       className={cn(
-                        "group flex min-h-8 min-w-0 max-w-full items-center gap-2 rounded-xl px-2 text-[13px] transition-colors",
+                        "group flex min-w-0 max-w-full items-center gap-2 rounded-xl px-2 text-[13px] transition-colors",
+                        compact ? "min-h-7" : "min-h-8",
                         active
                           ? "bg-sidebar-accent/70 text-sidebar-accent-foreground shadow-[inset_0_0_0_1px_hsl(var(--sidebar-border)/0.28)]"
                           : "text-sidebar-foreground/82 hover:bg-sidebar-accent/50 hover:text-sidebar-foreground",
@@ -79,11 +148,25 @@ export function ChatList({
                       <button
                         type="button"
                         onClick={() => onSelect(s.key)}
-                        title={rawLabel || fallbackTitle}
-                        className="min-w-0 flex-1 overflow-hidden py-1.5 text-left"
+                        title={tooltipTitle}
+                        className={cn(
+                          "min-w-0 flex-1 overflow-hidden text-left",
+                          compact ? "py-1" : "py-1.5",
+                        )}
                       >
                         <span className="block w-full truncate font-medium leading-5">{title}</span>
+                        {showPreview ? (
+                          <span className="block w-full truncate text-[11.5px] leading-4 text-muted-foreground/72">
+                            {preview}
+                          </span>
+                        ) : null}
+                        {timestamp ? (
+                          <span className="block w-full truncate text-[11px] leading-4 text-muted-foreground/58">
+                            {timestamp}
+                          </span>
+                        ) : null}
                       </button>
+                      <SessionActivityIndicator state={activityState} />
                       <DropdownMenu modal={false}>
                         <DropdownMenuTrigger
                           className={cn(
@@ -98,8 +181,35 @@ export function ChatList({
                         </DropdownMenuTrigger>
                         <DropdownMenuContent
                           align="end"
+                          portalContainer={actionMenuPortalContainer}
                           onCloseAutoFocus={(event) => event.preventDefault()}
                         >
+                          <DropdownMenuItem
+                            onSelect={() => onTogglePin(s.key)}
+                          >
+                            {isPinned ? (
+                              <PinOff className="mr-2 h-4 w-4" />
+                            ) : (
+                              <Pin className="mr-2 h-4 w-4" />
+                            )}
+                            {isPinned ? t("chat.unpin") : t("chat.pin")}
+                          </DropdownMenuItem>
+                          <DropdownMenuItem
+                            onSelect={() => onRequestRename(s.key, title)}
+                          >
+                            <Pencil className="mr-2 h-4 w-4" />
+                            {t("chat.rename")}
+                          </DropdownMenuItem>
+                          <DropdownMenuItem
+                            onSelect={() => onToggleArchive(s.key)}
+                          >
+                            {isArchived ? (
+                              <ArchiveRestore className="mr-2 h-4 w-4" />
+                            ) : (
+                              <Archive className="mr-2 h-4 w-4" />
+                            )}
+                            {isArchived ? t("chat.unarchive") : t("chat.archive")}
+                          </DropdownMenuItem>
                           <DropdownMenuItem
                             onSelect={() => {
                               window.setTimeout(() => onRequestDelete(s.key, title), 0);
@@ -123,16 +233,85 @@ export function ChatList({
   );
 }
 
+function SessionActivityIndicator({
+  state,
+}: {
+  state: "running" | "complete" | null;
+}) {
+  const { t } = useTranslation();
+
+  if (state === "running") {
+    const label = t("chat.activity.running");
+    return (
+      <span
+        aria-label={label}
+        title={label}
+        className="grid h-4 w-4 shrink-0 place-items-center"
+      >
+        <span className="h-3 w-3 animate-spin rounded-full border border-blue-500/25 border-t-blue-500 [animation-duration:1.4s] motion-reduce:animate-none dark:border-blue-400/25 dark:border-t-blue-400" />
+      </span>
+    );
+  }
+
+  if (state === "complete") {
+    const label = t("chat.activity.complete");
+    return (
+      <span
+        aria-label={label}
+        title={label}
+        className="grid h-4 w-4 shrink-0 place-items-center"
+      >
+        <span className="h-1.5 w-1.5 rounded-full bg-blue-500 shadow-[0_0_0_3px_rgba(59,130,246,0.14)] dark:bg-blue-400 dark:shadow-[0_0_0_3px_rgba(96,165,250,0.18)]" />
+      </span>
+    );
+  }
+
+  return <span className="h-4 w-4 shrink-0" aria-hidden="true" />;
+}
+
 function groupSessions(
   sessions: ChatSummary[],
-  labels: { today: string; yesterday: string; earlier: string },
+  labels: {
+    pinned: string;
+    all: string;
+    today: string;
+    yesterday: string;
+    earlier: string;
+    archived: string;
+    fallbackTitle: string;
+  },
+  options: {
+    pinnedKeys: string[];
+    archivedKeys: string[];
+    titleOverrides: Record<string, string>;
+    showArchived: boolean;
+    sort: SidebarSortMode;
+  },
 ): Array<{ label: string; sessions: ChatSummary[] }> {
   const now = new Date();
   const startOfToday = new Date(now.getFullYear(), now.getMonth(), now.getDate()).getTime();
   const startOfYesterday = startOfToday - 24 * 60 * 60 * 1000;
   const buckets = new Map<string, ChatSummary[]>();
+  const pinned = new Set(options.pinnedKeys);
+  const archived = new Set(options.archivedKeys);
+
+  const pinnedSessions: ChatSummary[] = [];
+  const archivedSessions: ChatSummary[] = [];
+  const normalSessions: ChatSummary[] = [];
 
   for (const session of sessions) {
+    if (archived.has(session.key)) {
+      if (options.showArchived) archivedSessions.push(session);
+      continue;
+    }
+    if (pinned.has(session.key)) {
+      pinnedSessions.push(session);
+      continue;
+    }
+    if (options.sort === "title_asc") {
+      normalSessions.push(session);
+      continue;
+    }
     const timestamp = Date.parse(session.updatedAt ?? session.createdAt ?? "");
     const label = Number.isFinite(timestamp) && timestamp >= startOfToday
       ? labels.today
@@ -144,7 +323,101 @@ function groupSessions(
     buckets.set(label, bucket);
   }
 
-  return [labels.today, labels.yesterday, labels.earlier]
-    .map((label) => ({ label, sessions: buckets.get(label) ?? [] }))
+  const groups = [labels.today, labels.yesterday, labels.earlier]
+    .map((label) => ({
+      label,
+      sessions: sortSessions(
+        buckets.get(label) ?? [],
+        options.sort,
+        options.titleOverrides,
+      ),
+    }))
     .filter((group) => group.sessions.length > 0);
+  if (options.sort === "title_asc" && normalSessions.length) {
+    groups.push({
+      label: labels.all,
+      sessions: sortSessions(
+        normalSessions,
+        options.sort,
+        options.titleOverrides,
+      ),
+    });
+  }
+  if (pinnedSessions.length) {
+    groups.unshift({
+      label: labels.pinned,
+      sessions: sortSessions(
+        pinnedSessions,
+        options.sort,
+        options.titleOverrides,
+      ),
+    });
+  }
+  if (archivedSessions.length) {
+    groups.push({
+      label: labels.archived,
+      sessions: sortSessions(
+        archivedSessions,
+        options.sort,
+        options.titleOverrides,
+      ),
+    });
+  }
+  return groups;
+}
+
+function sortSessions(
+  sessions: ChatSummary[],
+  sort: SidebarSortMode,
+  titleOverrides: Record<string, string>,
+): ChatSummary[] {
+  const copy = [...sessions];
+  copy.sort((a, b) => {
+    if (sort === "title_asc") {
+      const titleOrder = titleForSort(a, titleOverrides).localeCompare(
+        titleForSort(b, titleOverrides),
+        "en",
+        { numeric: true, sensitivity: "base" },
+      );
+      if (titleOrder !== 0) return titleOrder;
+      return sessionTime(b, "updatedAt") - sessionTime(a, "updatedAt");
+    }
+    const aTime = sessionTime(a, sort === "created_desc" ? "createdAt" : "updatedAt");
+    const bTime = sessionTime(b, sort === "created_desc" ? "createdAt" : "updatedAt");
+    return bTime - aTime;
+  });
+  return copy;
+}
+
+function titleForSort(
+  session: ChatSummary,
+  titleOverrides: Record<string, string>,
+): string {
+  return (
+    titleOverrides[session.key]?.trim() ||
+    session.title?.trim() ||
+    deriveTitle(session.preview, "new chat")
+  ).toLocaleLowerCase("en");
+}
+
+function displayTitle(
+  session: ChatSummary,
+  titleOverrides: Record<string, string>,
+  fallbackTitle: string,
+): string {
+  return (
+    titleOverrides[session.key]?.trim() ||
+    session.title?.trim() ||
+    deriveTitle(session.preview, fallbackTitle)
+  );
+}
+
+function sessionTime(
+  session: ChatSummary,
+  field: "createdAt" | "updatedAt",
+): number {
+  const primary = Date.parse(session[field] ?? "");
+  if (Number.isFinite(primary)) return primary;
+  const fallback = Date.parse(session.updatedAt ?? session.createdAt ?? "");
+  return Number.isFinite(fallback) ? fallback : 0;
 }

webui/src/components/CodeBlock.tsx

@@ -1,44 +1,75 @@
-import { useCallback, useEffect, useState } from "react";
+import { Suspense, lazy, useCallback, useState } from "react";
 import { Check, Copy } from "lucide-react";
 import { useTranslation } from "react-i18next";
-import { Prism as SyntaxHighlighter } from "react-syntax-highlighter";
-import {
-  oneDark,
-  oneLight,
-} from "react-syntax-highlighter/dist/esm/styles/prism";
 
+import { useThemeValue } from "@/hooks/useTheme";
 import { cn } from "@/lib/utils";
 
 interface CodeBlockProps {
   language?: string;
   code: string;
   className?: string;
+  highlight?: boolean;
 }
 
-/** Read dark mode straight from the DOM — stays in sync with Tailwind's `dark:`. */
-function useIsDark() {
-  const [isDark, setIsDark] = useState(() =>
-    typeof document !== "undefined"
-      ? document.documentElement.classList.contains("dark")
-      : true,
+interface HighlightedCodeProps {
+  language?: string;
+  code: string;
+  isDark: boolean;
+}
+
+const LazyHighlightedCode = lazy(async () => {
+  const [
+    { default: SyntaxHighlighter },
+    { default: oneDark },
+    { default: oneLight },
+  ] = await Promise.all([
+    import("react-syntax-highlighter/dist/esm/prism-async-light"),
+    import("react-syntax-highlighter/dist/esm/styles/prism/one-dark"),
+    import("react-syntax-highlighter/dist/esm/styles/prism/one-light"),
+  ]);
+
+  return {
+    default({ language, code, isDark }: HighlightedCodeProps) {
+      return (
+        <SyntaxHighlighter
+          language={language}
+          style={isDark ? oneDark : oneLight}
+          customStyle={{
+            margin: 0,
+            padding: "1rem",
+            fontSize: "0.875rem",
+            lineHeight: 1.6,
+          }}
+          PreTag="pre"
+          wrapLongLines
+        >
+          {code}
+        </SyntaxHighlighter>
+      );
+    },
+  };
+});
+
+function PlainCodeFallback({ code }: { code: string }) {
+  return (
+    <pre
+      className="m-0 overflow-x-auto whitespace-pre-wrap p-4 font-mono text-sm leading-[1.6]"
+    >
+      <code>{code}</code>
+    </pre>
   );
-
-  useEffect(() => {
-    const el = document.documentElement;
-    const observer = new MutationObserver(() => {
-      setIsDark(el.classList.contains("dark"));
-    });
-    observer.observe(el, { attributeFilter: ["class"] });
-    return () => observer.disconnect();
-  }, []);
-
-  return isDark;
 }
 
-export function CodeBlock({ language, code, className }: CodeBlockProps) {
+export function CodeBlock({
+  language,
+  code,
+  className,
+  highlight = true,
+}: CodeBlockProps) {
   const { t } = useTranslation();
   const [copied, setCopied] = useState(false);
-  const isDark = useIsDark();
+  const isDark = useThemeValue() === "dark";
 
   const onCopy = useCallback(() => {
     if (!navigator.clipboard) return;
@@ -86,20 +117,13 @@ export function CodeBlock({ language, code, className }: CodeBlockProps) {
           <span>{copied ? t("code.copied") : t("code.copy")}</span>
         </button>
       </div>
-      <SyntaxHighlighter
-        language={language}
-        style={isDark ? oneDark : oneLight}
-        customStyle={{
-          margin: 0,
-          padding: "1rem",
-          fontSize: "0.875rem",
-          lineHeight: 1.6,
-        }}
-        PreTag="pre"
-        wrapLongLines
-      >
-        {code}
-      </SyntaxHighlighter>
+      {highlight ? (
+        <Suspense fallback={<PlainCodeFallback code={code} />}>
+          <LazyHighlightedCode language={language} code={code} isDark={isDark} />
+        </Suspense>
+      ) : (
+        <PlainCodeFallback code={code} />
+      )}
     </div>
   );
 }

webui/src/components/ConnectionBadge.tsx

@@ -36,21 +36,25 @@ export function ConnectionBadge() {
     status === "connecting" ||
     status === "reconnecting" ||
     status === "error";
+  const label = t(`connection.${status}`);
   return (
     <span
       className={cn(
-        "inline-flex min-w-0 items-center gap-1.5 rounded-md px-1.5 py-1 text-[11px] font-medium transition-colors",
+        "inline-flex h-8 w-8 shrink-0 items-center justify-center rounded-full transition-colors",
+        "text-muted-foreground/70 hover:bg-sidebar-accent/65",
         meta.color,
       )}
       aria-live="polite"
+      role="status"
+      title={label}
     >
-      <span className="relative flex h-1.5 w-1.5" aria-hidden>
+      <span className="relative flex h-2 w-2" aria-hidden>
         {pulsing && (
           <span className="absolute inline-flex h-full w-full animate-ping rounded-full bg-current opacity-75" />
         )}
-        <span className="relative inline-flex h-1.5 w-1.5 rounded-full bg-current" />
+        <span className="relative inline-flex h-2 w-2 rounded-full bg-current" />
       </span>
-      {t(`connection.${status}`)}
+      <span className="sr-only">{label}</span>
     </span>
   );
 }

webui/src/components/FileReferenceChip.tsx

@@ -0,0 +1,230 @@
+import {
+  Tooltip,
+  TooltipContent,
+  TooltipProvider,
+  TooltipTrigger,
+} from "@/components/ui/tooltip";
+import { cn } from "@/lib/utils";
+
+type FileReferenceKind =
+  | "default"
+  | "css"
+  | "html"
+  | "json"
+  | "markdown"
+  | "notebook"
+  | "python"
+  | "react"
+  | "typescript";
+
+interface FileReferenceChipProps {
+  path: string;
+  tooltipPath?: string;
+  display?: "name" | "path";
+  active?: boolean;
+  className?: string;
+  textClassName?: string;
+  testId?: string;
+}
+
+export function FileReferenceChip({
+  path,
+  tooltipPath,
+  display = "name",
+  active = false,
+  className,
+  textClassName,
+  testId = "inline-file-path",
+}: FileReferenceChipProps) {
+  const { directory, name } = splitFilePath(path);
+  const kind = fileKindForPath(path);
+  const displayText = display === "path" ? path.replace(/\\/g, "/") : name;
+  const fullPath = tooltipPath || path;
+  return (
+    <TooltipProvider delayDuration={500} skipDelayDuration={100}>
+      <Tooltip>
+        <TooltipTrigger asChild>
+          <span
+            className={cn("not-prose inline-flex max-w-full align-baseline leading-[inherit]", className)}
+          >
+            <span
+              data-testid={testId}
+              aria-label={fullPath}
+              className={cn(
+                "inline-flex max-w-full items-center gap-1 font-medium leading-[inherit]",
+                "text-sky-600 transition-colors hover:text-sky-700",
+                "dark:text-sky-300 dark:hover:text-sky-200",
+              )}
+            >
+              <FileReferenceIcon kind={kind} />
+              <span
+                data-sheen-text={active ? displayText : undefined}
+                className={cn(
+                  "min-w-0 max-w-full truncate",
+                  active && "streaming-text-sheen file-reference-sheen",
+                  textClassName,
+                )}
+              >
+                {display === "path" && directory ? (
+                  <>
+                    <span className="text-muted-foreground/65">{directory}</span>
+                    <span className="font-semibold text-sky-700 dark:text-sky-200">{name}</span>
+                  </>
+                ) : (
+                  displayText
+                )}
+              </span>
+            </span>
+          </span>
+        </TooltipTrigger>
+        <TooltipContent
+          side="top"
+          align="center"
+          sideOffset={8}
+          collisionPadding={12}
+          className={cn(
+            "max-w-[min(38rem,calc(100vw-2rem))] rounded-[10px]",
+            "border-border/60 bg-popover/95 px-2.5 py-1.5",
+            "break-all font-mono text-[11px] leading-snug text-popover-foreground",
+            "shadow-lg backdrop-blur",
+          )}
+        >
+          {fullPath}
+        </TooltipContent>
+      </Tooltip>
+    </TooltipProvider>
+  );
+}
+
+export function isLikelyFilePath(value: string): boolean {
+  const raw = value.trim();
+  if (!raw || raw.includes("\n")) return false;
+  if (/^[a-z][a-z0-9+.-]*:\/\//i.test(raw)) return false;
+  if (!/[\\/]/.test(raw) && !/^(dockerfile|makefile|readme|package-lock\.json)$/i.test(raw)) {
+    return false;
+  }
+  const normalized = raw.replace(/\\/g, "/");
+  const name = normalized.split("/").filter(Boolean).pop() ?? normalized;
+  if (!name || name === "." || name === "..") return false;
+  if (/^(dockerfile|makefile|readme|package-lock\.json)$/i.test(name)) return true;
+  return /\.[a-z0-9][a-z0-9_-]{0,12}$/i.test(name);
+}
+
+function splitFilePath(path: string): { directory: string; name: string } {
+  const normalized = path.replace(/\\/g, "/");
+  const slash = normalized.lastIndexOf("/");
+  if (slash < 0) return { directory: "", name: path };
+  return {
+    directory: normalized.slice(0, slash + 1),
+    name: normalized.slice(slash + 1) || normalized,
+  };
+}
+
+function fileKindForPath(path: string): FileReferenceKind {
+  const normalized = path.toLowerCase();
+  const name = normalized.split(/[\\/]/).pop() ?? normalized;
+  const ext = name.includes(".") ? name.split(".").pop() ?? "" : "";
+  if (name === "dockerfile") {
+    return "default";
+  }
+  switch (ext) {
+    case "py":
+    case "pyi":
+      return "python";
+    case "jsx":
+    case "tsx":
+      return "react";
+    case "ts":
+      return "typescript";
+    case "html":
+    case "htm":
+      return "html";
+    case "css":
+    case "scss":
+    case "sass":
+      return "css";
+    case "json":
+    case "jsonl":
+      return "json";
+    case "md":
+    case "mdx":
+      return "markdown";
+    case "ipynb":
+      return "notebook";
+    default:
+      return "default";
+  }
+}
+
+function FileReferenceIcon({ kind }: { kind: FileReferenceKind }) {
+  if (kind === "react") {
+    return (
+      <svg
+        aria-hidden
+        className="h-[0.98em] w-[0.98em] shrink-0 text-sky-500 dark:text-sky-300"
+        viewBox="0 0 24 24"
+        fill="none"
+        stroke="currentColor"
+        strokeWidth="1.6"
+        strokeLinecap="round"
+        strokeLinejoin="round"
+      >
+        <circle cx="12" cy="12" r="1.9" fill="currentColor" stroke="none" />
+        <ellipse cx="12" cy="12" rx="9" ry="3.7" />
+        <ellipse cx="12" cy="12" rx="9" ry="3.7" transform="rotate(60 12 12)" />
+        <ellipse cx="12" cy="12" rx="9" ry="3.7" transform="rotate(120 12 12)" />
+      </svg>
+    );
+  }
+  if (kind === "default") {
+    return (
+      <svg
+        aria-hidden
+        className="h-[0.98em] w-[0.98em] shrink-0 text-sky-500 dark:text-sky-300"
+        viewBox="0 0 24 24"
+        fill="none"
+        stroke="currentColor"
+        strokeWidth="1.9"
+        strokeLinecap="round"
+        strokeLinejoin="round"
+      >
+        <path d="M14 2H7a2 2 0 0 0-2 2v16a2 2 0 0 0 2 2h10a2 2 0 0 0 2-2V7z" />
+        <path d="M14 2v5h5" />
+      </svg>
+    );
+  }
+  const label = fileKindLabel(kind);
+  return (
+    <span
+      aria-hidden
+      className={cn(
+        "inline-flex h-[1.05em] min-w-[1.05em] shrink-0 items-center justify-center",
+        "rounded-[4px] bg-sky-500/12 px-[0.22em] text-[0.58em] font-bold uppercase leading-none",
+        "text-sky-600 dark:bg-sky-400/15 dark:text-sky-300",
+      )}
+    >
+      {label}
+    </span>
+  );
+}
+
+function fileKindLabel(kind: FileReferenceKind): string {
+  switch (kind) {
+    case "css":
+      return "#";
+    case "html":
+      return "H";
+    case "json":
+      return "{}";
+    case "markdown":
+      return "M";
+    case "notebook":
+      return "N";
+    case "python":
+      return "PY";
+    case "typescript":
+      return "TS";
+    default:
+      return "";
+  }
+}