Merge remote-tracking branch 'origin/main' into nightly

README.md

@@ -23,6 +23,14 @@
 
 ## 📢 News
 
+- **2026-04-29** 🚀 Released **v0.1.5.post3** — Smarter threads on Feishu, Discord, Slack, and Teams; **DeepSeek-V4**; Hugging Face & Olostep; choices, `/history`, and steadier long chats. Please see [release notes](https://github.com/HKUDS/nanobot/releases/tag/v0.1.5.post3) for details.
+- **2026-04-28** 🌐 Olostep web search, Hugging Face provider, safer workspace-tool interruptions.
+- **2026-04-27** 💬 `/history` command, smarter session replay caps, smoother Discord / Slack threads.
+- **2026-04-26** 🧭 Natural cron reminders, thread-aware restarts, safer local provider and shell behavior.
+- **2026-04-25** 🧩 `ask_user` choices, macOS LaunchAgent deployment, MSTeams stale-reference cleanup.
+- **2026-04-24** 🎥 Video attachments for channels, DeepSeek thinking control, faster document startup.
+- **2026-04-23** 🧵 Discord thread sessions, Telegram inline buttons, structured tool progress updates.
+- **2026-04-22** 🔎 GitHub Copilot GPT-5 / o-series support, configurable web fetch, WebUI image uploads.
 - **2026-04-21** 🚀 Released **v0.1.5.post2** — Windows & Python 3.14 support, Office document reading, SSE streaming for the OpenAI-compatible API, and stronger reliability across sessions, memory, and channels. Please see [release notes](https://github.com/HKUDS/nanobot/releases/tag/v0.1.5.post2) for details.
 - **2026-04-20** 🎨 Kimi K2.6 support, Telegram long-message split, WebUI typography & dark-mode polish.
 - **2026-04-19** 🌐 WebUI i18n locale switcher, atomic session writes with auto-repair.

bridge/src/whatsapp.ts

@@ -17,7 +17,7 @@ import { Boom } from '@hapi/boom';
 import qrcode from 'qrcode-terminal';
 import pino from 'pino';
 import { readFile, writeFile, mkdir } from 'fs/promises';
-import { join, basename } from 'path';
+import { join, basename, resolve, sep } from 'path';
 import { randomBytes } from 'crypto';
 
 const VERSION = '0.1.0';
@@ -196,17 +196,18 @@ export class WhatsAppClient {
 
       let outFilename: string;
       if (fileName) {
-        // Documents have a filename — use it with a unique prefix to avoid collisions
+        const safeName = basename(fileName).replace(/[^a-zA-Z0-9._-]/g, '_');
-        const prefix = `wa_${Date.now()}_${randomBytes(4).toString('hex')}_`;
+        outFilename = `wa_${Date.now()}_${randomBytes(4).toString('hex')}_${safeName}`;
-        outFilename = prefix + fileName;
       } else {
         const mime = mimetype || 'application/octet-stream';
-        // Derive extension from mimetype subtype (e.g. "image/png" → ".png", "application/pdf" → ".pdf")
         const ext = '.' + (mime.split('/').pop()?.split(';')[0] || 'bin');
         outFilename = `wa_${Date.now()}_${randomBytes(4).toString('hex')}${ext}`;
       }
 
-      const filepath = join(mediaDir, outFilename);
+      const filepath = resolve(mediaDir, outFilename);
+      if (!filepath.startsWith(resolve(mediaDir) + sep)) {
+        throw new Error(`Path traversal blocked: ${outFilename}`);
+      }
       await writeFile(filepath, buffer);
 
       return filepath;

docs/configuration.md

@@ -474,6 +474,26 @@ Global settings that apply to all channels. Configure under the `channels` secti
 | `transcriptionProvider` | `"groq"` | Voice transcription backend: `"groq"` (free tier, default) or `"openai"`. API key is auto-resolved from the matching provider config. |
 | `transcriptionLanguage` | `null` | Optional ISO-639-1 language hint for audio transcription, e.g. `"en"`, `"ko"`, `"ja"`. |
 
+`sendProgress` and `sendToolHints` can also be overridden per channel. The
+global values stay as defaults for channels that do not set their own value:
+
+```json
+{
+  "channels": {
+    "sendProgress": true,
+    "sendToolHints": false,
+    "telegram": {
+      "enabled": true,
+      "sendProgress": false
+    },
+    "websocket": {
+      "enabled": true,
+      "sendToolHints": true
+    }
+  }
+}
+```
+
 ### Retry Behavior
 
 Retry is intentionally simple.

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.post2"
+        return _read_pyproject_version() or "0.1.5.post3"
 
 
 __version__ = _resolve_version()

nanobot/agent/loop.py

@@ -258,6 +258,7 @@ class AgentLoop:
             exec_config=self.exec_config,
             restrict_to_workspace=restrict_to_workspace,
             disabled_skills=disabled_skills,
+            max_iterations=self.max_iterations,
         )
         self._unified_session = unified_session
         self._max_messages = max_messages if max_messages > 0 else 120
@@ -307,6 +308,10 @@ class AgentLoop:
         self.commands = CommandRouter()
         register_builtin_commands(self.commands)
 
+    def _sync_subagent_runtime_limits(self) -> None:
+        """Keep subagent runtime limits aligned with mutable loop settings."""
+        self.subagents.max_iterations = self.max_iterations
+
     def _apply_provider_snapshot(self, snapshot: ProviderSnapshot) -> None:
         """Swap model/provider for future turns without disturbing an active one."""
         provider = snapshot.provider
@@ -531,6 +536,8 @@ class AgentLoop:
 
         Returns (final_content, tools_used, messages, stop_reason, had_injections).
         """
+        self._sync_subagent_runtime_limits()
+
         loop_hook = _LoopHook(
             self,
             on_progress=on_progress,

nanobot/agent/memory.py

@@ -4,6 +4,7 @@ from __future__ import annotations
 
 import asyncio
 import json
+import os
 import re
 import weakref
 import tiktoken
@@ -359,10 +360,31 @@ class MemoryStore:
             return None
 
     def _write_entries(self, entries: list[dict[str, Any]]) -> None:
-        """Overwrite history.jsonl with the given entries."""
+        """Overwrite history.jsonl with the given entries (atomic write)."""
-        with open(self.history_file, "w", encoding="utf-8") as f:
+        tmp_path = self.history_file.with_suffix(self.history_file.suffix + ".tmp")
-            for entry in entries:
+        try:
-                f.write(json.dumps(entry, ensure_ascii=False) + "\n")
+            with open(tmp_path, "w", encoding="utf-8") as f:
+                for entry in entries:
+                    f.write(json.dumps(entry, ensure_ascii=False) + "\n")
+                f.flush()
+                os.fsync(f.fileno())
+            os.replace(tmp_path, self.history_file)
+
+            # fsync the directory so the rename is durable.
+            # On Windows, opening a directory with O_RDONLY raises
+            # PermissionError — skip the dir sync there (NTFS
+            # journals metadata synchronously).
+            try:
+                fd = os.open(str(self.history_file.parent), os.O_RDONLY)
+                try:
+                    os.fsync(fd)
+                finally:
+                    os.close(fd)
+            except PermissionError:
+                pass  # Windows — directory fsync not supported
+        except BaseException:
+            tmp_path.unlink(missing_ok=True)
+            raise
 
     # -- dream cursor --------------------------------------------------------
 

nanobot/agent/subagent.py

@@ -20,7 +20,7 @@ from nanobot.agent.tools.shell import ExecTool
 from nanobot.agent.tools.web import WebFetchTool, WebSearchTool
 from nanobot.bus.events import InboundMessage
 from nanobot.bus.queue import MessageBus
-from nanobot.config.schema import ExecToolConfig, WebToolsConfig
+from nanobot.config.schema import AgentDefaults, ExecToolConfig, WebToolsConfig
 from nanobot.providers.base import LLMProvider
 from nanobot.utils.prompt_templates import render_template
 
@@ -81,6 +81,7 @@ class SubagentManager:
         exec_config: "ExecToolConfig | None" = None,
         restrict_to_workspace: bool = False,
         disabled_skills: list[str] | None = None,
+        max_iterations: int | None = None,
     ):
         self.provider = provider
         self.workspace = workspace
@@ -91,6 +92,11 @@ class SubagentManager:
         self.exec_config = exec_config or ExecToolConfig()
         self.restrict_to_workspace = restrict_to_workspace
         self.disabled_skills = set(disabled_skills or [])
+        self.max_iterations = (
+            max_iterations
+            if max_iterations is not None
+            else AgentDefaults().max_tool_iterations
+        )
         self.runner = AgentRunner(provider)
         self._running_tasks: dict[str, asyncio.Task[None]] = {}
         self._task_statuses: dict[str, SubagentStatus] = {}
@@ -202,7 +208,7 @@ class SubagentManager:
                 initial_messages=messages,
                 tools=tools,
                 model=self.model,
-                max_iterations=15,
+                max_iterations=self.max_iterations,
                 max_tool_result_chars=self.max_tool_result_chars,
                 hook=_SubagentHook(task_id, status),
                 max_iterations_message="Task completed but no final response was generated.",

nanobot/agent/tools/self.py

@@ -394,6 +394,8 @@ class MyTool(Tool):
         if "min_len" in spec and len(str(value)) < spec["min_len"]:
             return f"Error: '{key}' must be at least {spec['min_len']} characters"
         setattr(self._loop, key, value)
+        if key == "max_iterations" and hasattr(self._loop, "_sync_subagent_runtime_limits"):
+            self._loop._sync_subagent_runtime_limits()
         self._audit("modify", f"{key}: {old!r} -> {value!r}")
         return f"Set {key} = {value!r} (was {old!r})"
 

nanobot/channels/base.py

@@ -26,6 +26,8 @@ class BaseChannel(ABC):
     transcription_api_key: str = ""
     transcription_api_base: str = ""
     transcription_language: str | None = None
+    send_progress: bool = True
+    send_tool_hints: bool = False
 
     def __init__(self, config: Any, bus: MessageBus):
         """

nanobot/channels/feishu.py

@@ -1170,6 +1170,21 @@ class FeishuChannel(BaseChannel):
             logger.error("Error replying to Feishu message {}: {}", parent_message_id, e)
             return False
 
+    def _should_use_reply_in_thread(self, metadata: dict[str, Any]) -> bool:
+        """Return whether a group reply should create a Feishu thread/topic."""
+        return metadata.get("chat_type", "group") == "group" and self.config.reply_to_message
+
+    def _thread_reply_target(self, metadata: dict[str, Any]) -> str | None:
+        """Return the message_id that should receive a Reply API response."""
+        if metadata.get("chat_type", "group") != "group":
+            return None
+        message_id = metadata.get("message_id")
+        if not message_id:
+            return None
+        if metadata.get("thread_id") or self.config.reply_to_message:
+            return message_id
+        return None
+
     def _send_message_sync(
         self, receive_id_type: str, receive_id: str, msg_type: str, content: str
     ) -> str | None:
@@ -1211,13 +1226,15 @@ class FeishuChannel(BaseChannel):
         receive_id_type: str,
         chat_id: str,
         reply_message_id: str | None = None,
+        *,
+        reply_in_thread: bool = False,
     ) -> str | None:
         """Create a CardKit streaming card, send it to chat, return card_id.
 
         When *reply_message_id* is provided the card is delivered via the
-        reply API (with reply_in_thread=True) so it lands inside the
+        reply API. *reply_in_thread* controls whether Feishu creates a
-        originating thread / topic.  Otherwise the plain create-message
+        thread/topic for that reply. Otherwise the plain create-message API is
-        API is used.
+        used.
         """
         from lark_oapi.api.cardkit.v1 import CreateCardRequest, CreateCardRequestBody
 
@@ -1253,7 +1270,7 @@ class FeishuChannel(BaseChannel):
                 if reply_message_id:
                     sent = self._reply_message_sync(
                         reply_message_id, "interactive", card_content,
-                        reply_in_thread=True,
+                        reply_in_thread=reply_in_thread,
                     )
                 else:
                     sent = self._send_message_sync(
@@ -1409,16 +1426,14 @@ class FeishuChannel(BaseChannel):
                     {"config": {"wide_screen_mode": True}, "elements": chunk},
                     ensure_ascii=False,
                 )
-                # Fallback: reply via the Reply API for group chats.
+                # Fallback replies stay in existing topics, but only create a
-                # Target message_id — the Feishu API keeps the reply in
+                # new topic when reply-to-message is enabled.
-                # the same topic automatically.
+                fallback_msg_id = self._thread_reply_target(meta)
-                _f_msg = meta.get("message_id")
-                fallback_msg_id = _f_msg if meta.get("chat_type", "group") == "group" else None
                 if fallback_msg_id:
                     await loop.run_in_executor(
                         None, lambda: self._reply_message_sync(
                             fallback_msg_id, "interactive", card,
-                            reply_in_thread=True,
+                            reply_in_thread=self._should_use_reply_in_thread(meta),
                         ),
                     )
                 else:
@@ -1438,16 +1453,18 @@ class FeishuChannel(BaseChannel):
 
         now = time.monotonic()
         if buf.card_id is None:
-            # Send the streaming card as a reply for group chats so it
+            # Use the Reply API for existing topics, and only create new topics
-            # lands inside the originating topic/thread.  Always target
+            # when reply-to-message is enabled.
-            # message_id (the actual inbound message) — the Feishu Reply
+            use_reply_in_thread = self._should_use_reply_in_thread(meta)
-            # API keeps the response in the same topic automatically.
+            reply_msg_id = self._thread_reply_target(meta)
-            is_group = meta.get("chat_type", "group") == "group"
-            reply_msg_id = meta.get("message_id") if is_group else None
             card_id = await loop.run_in_executor(
                 None,
-                self._create_streaming_card_sync,
+                lambda: self._create_streaming_card_sync(
-                rid_type, chat_id, reply_msg_id,
+                    rid_type,
+                    chat_id,
+                    reply_msg_id,
+                    reply_in_thread=use_reply_in_thread,
+                ),
             )
             if card_id:
                 buf.card_id = card_id
@@ -1489,22 +1506,21 @@ class FeishuChannel(BaseChannel):
                         "\n\n" + self._format_tool_hint_delta(hint) + "\n\n",
                     )
                     return
-                # No active streaming card — send as a regular
+                # No active streaming card — send as a regular interactive card
-                # interactive card with the same 🔧 prefix style.
+                # with the same 🔧 prefix style. Existing topics stay threaded;
-                # Use reply API for group chats so the hint stays in topic.
+                # new topics are created only when reply-to-message is enabled.
                 card = json.dumps(
                     {"config": {"wide_screen_mode": True}, "elements": [
                         {"tag": "markdown", "content": self._format_tool_hint_delta(hint)},
                     ]},
                     ensure_ascii=False,
                 )
-                _th_msg_id = msg.metadata.get("message_id")
+                _th_msg_id = self._thread_reply_target(msg.metadata)
-                _th_chat_type = msg.metadata.get("chat_type", "group")
+                if _th_msg_id:
-                if _th_msg_id and _th_chat_type == "group":
                     await loop.run_in_executor(
                         None, lambda: self._reply_message_sync(
                             _th_msg_id, "interactive", card,
-                            reply_in_thread=True,
+                            reply_in_thread=self._should_use_reply_in_thread(msg.metadata),
                         ),
                     )
                 else:
@@ -1531,18 +1547,16 @@ class FeishuChannel(BaseChannel):
             def _do_send(m_type: str, content: str) -> None:
                 """Send via reply (first message) or create (subsequent).
 
-                For group chats the reply API always uses reply_in_thread=True.
+                Group chats only set reply_in_thread=True when
-                The Feishu API automatically keeps replies inside existing
+                reply_to_message is enabled; otherwise a Reply API call for an
-                topics — reply_in_thread only creates a *new* topic when the
+                existing topic must not create a new topic.
-                target message is a plain (non-topic) message.
                 """
                 nonlocal first_send
                 if reply_message_id and first_send:
                     first_send = False
-                    chat_type = msg.metadata.get("chat_type", "group")
                     ok = self._reply_message_sync(
                         reply_message_id, m_type, content,
-                        reply_in_thread=chat_type == "group",
+                        reply_in_thread=self._should_use_reply_in_thread(msg.metadata),
                     )
                     if ok:
                         return

nanobot/channels/manager.py

@@ -27,9 +27,15 @@ def _default_webui_dist() -> Path | None:
     candidate = Path(web_pkg.__file__).resolve().parent / "dist"
     return candidate if candidate.is_dir() else None
 
+
 # Retry delays for message sending (exponential backoff: 1s, 2s, 4s)
 _SEND_RETRY_DELAYS = (1, 2, 4)
 
+_BOOL_CAMEL_ALIASES: dict[str, str] = {
+    "send_progress": "sendProgress",
+    "send_tool_hints": "sendToolHints",
+}
+
 
 class ChannelManager:
     """
@@ -90,6 +96,12 @@ class ChannelManager:
                 channel.transcription_api_key = transcription_key
                 channel.transcription_api_base = transcription_base
                 channel.transcription_language = transcription_language
+                channel.send_progress = self._resolve_bool_override(
+                    section, "send_progress", self.config.channels.send_progress,
+                )
+                channel.send_tool_hints = self._resolve_bool_override(
+                    section, "send_tool_hints", self.config.channels.send_tool_hints,
+                )
                 self.channels[name] = channel
                 logger.info("{} channel enabled", cls.display_name)
             except Exception as e:
@@ -131,6 +143,31 @@ class ChannelManager:
                     f'Set ["*"] to allow everyone, or add specific user IDs.'
                 )
 
+    def _should_send_progress(self, channel_name: str, *, tool_hint: bool = False) -> bool:
+        """Return whether progress (or tool-hints) may be sent to *channel_name*."""
+        ch = self.channels.get(channel_name)
+        if ch is None:
+            logger.warning("Progress check for unknown channel: {}", channel_name)
+            return False
+        return ch.send_tool_hints if tool_hint else ch.send_progress
+
+    def _resolve_bool_override(self, section: Any, key: str, default: bool) -> bool:
+        """Return *key* from *section* if it is a bool, otherwise *default*.
+
+        For dict configs also checks the camelCase alias (e.g. ``sendProgress``
+        for ``send_progress``) so raw JSON/TOML configs work alongside
+        Pydantic models.
+        """
+        if isinstance(section, dict):
+            value = section.get(key)
+            if value is None:
+                camel = _BOOL_CAMEL_ALIASES.get(key)
+                if camel:
+                    value = section.get(camel)
+            return value if isinstance(value, bool) else default
+        value = getattr(section, key, None)
+        return value if isinstance(value, bool) else default
+
     async def _start_channel(self, name: str, channel: BaseChannel) -> None:
         """Start a channel and log any exceptions."""
         try:
@@ -216,9 +253,13 @@ class ChannelManager:
                     )
 
                 if msg.metadata.get("_progress"):
-                    if msg.metadata.get("_tool_hint") and not self.config.channels.send_tool_hints:
+                    if msg.metadata.get("_tool_hint") and not self._should_send_progress(
+                        msg.channel, tool_hint=True,
+                    ):
                         continue
-                    if not msg.metadata.get("_tool_hint") and not self.config.channels.send_progress:
+                    if not msg.metadata.get("_tool_hint") and not self._should_send_progress(
+                        msg.channel, tool_hint=False,
+                    ):
                         continue
 
                 if msg.metadata.get("_retry_wait"):

nanobot/channels/matrix.py

@@ -262,10 +262,18 @@ class MatrixChannel(BaseChannel):
         self.store_path.mkdir(parents=True, exist_ok=True)
         self.session_path = self.store_path / "session.json"
 
+        # Replace ':' with '_' to produce a Windows-safe filename
+        safe_store_name = self.config.user_id.replace(":", "_") + f"_{self.config.device_id}.db"
+
         self.client = AsyncClient(
-            homeserver=self.config.homeserver, user=self.config.user_id,
+            homeserver=self.config.homeserver,
+            user=self.config.user_id,
             store_path=self.store_path,
-            config=AsyncClientConfig(store_sync_tokens=True, encryption_enabled=self.config.e2ee_enabled),
+            config=AsyncClientConfig(
+                store_sync_tokens=True,
+                encryption_enabled=self.config.e2ee_enabled,
+                store_name=safe_store_name,
+            ),
         )
 
         self._register_event_callbacks()

nanobot/channels/whatsapp.py

@@ -1,6 +1,7 @@
 """WhatsApp channel implementation using Node.js bridge."""
 
 import asyncio
+import hashlib
 import json
 import mimetypes
 import os
@@ -316,13 +317,7 @@ def _ensure_bridge_setup() -> Path:
     from nanobot.config.paths import get_bridge_install_dir
 
     user_bridge = get_bridge_install_dir()
-
+    stamp_file = user_bridge / ".nanobot-bridge-source-hash"
-    if (user_bridge / "dist" / "index.js").exists():
-        return user_bridge
-
-    npm_path = shutil.which("npm")
-    if not npm_path:
-        raise RuntimeError("npm not found. Please install Node.js >= 18.")
 
     # Find source bridge
     current_file = Path(__file__)
@@ -341,6 +336,33 @@ def _ensure_bridge_setup() -> Path:
             "Try reinstalling: pip install --force-reinstall nanobot"
         )
 
+    def source_hash(root: Path) -> str:
+        digest = hashlib.sha256()
+        for path in sorted(root.rglob("*")):
+            if not path.is_file():
+                continue
+            rel = path.relative_to(root)
+            if rel.parts and rel.parts[0] in {"node_modules", "dist"}:
+                continue
+            digest.update(rel.as_posix().encode("utf-8"))
+            digest.update(b"\0")
+            digest.update(path.read_bytes())
+            digest.update(b"\0")
+        return digest.hexdigest()
+
+    expected_hash = source_hash(source)
+    current_hash = stamp_file.read_text().strip() if stamp_file.exists() else None
+
+    if (user_bridge / "dist" / "index.js").exists() and current_hash == expected_hash:
+        return user_bridge
+
+    if (user_bridge / "dist" / "index.js").exists() and current_hash != expected_hash:
+        logger.info("WhatsApp bridge source changed; rebuilding bridge...")
+
+    npm_path = shutil.which("npm")
+    if not npm_path:
+        raise RuntimeError("npm not found. Please install Node.js >= 18.")
+
     logger.info("Setting up WhatsApp bridge...")
     user_bridge.parent.mkdir(parents=True, exist_ok=True)
     if user_bridge.exists():
@@ -352,6 +374,7 @@ def _ensure_bridge_setup() -> Path:
 
     logger.info("  Building...")
     subprocess.run([npm_path, "run", "build"], cwd=user_bridge, check=True, capture_output=True)
+    stamp_file.write_text(expected_hash + "\n")
 
     logger.info("Bridge ready")
     return user_bridge

nanobot/cli/commands.py

@@ -1271,6 +1271,7 @@ def channels_status(
 
 def _get_bridge_dir() -> Path:
     """Get the bridge directory, setting it up if needed."""
+    import hashlib
     import shutil
     import subprocess
 
@@ -1278,16 +1279,7 @@ def _get_bridge_dir() -> Path:
     from nanobot.config.paths import get_bridge_install_dir
 
     user_bridge = get_bridge_install_dir()
-
+    stamp_file = user_bridge / ".nanobot-bridge-source-hash"
-    # Check if already built
-    if (user_bridge / "dist" / "index.js").exists():
-        return user_bridge
-
-    # Check for npm
-    npm_path = shutil.which("npm")
-    if not npm_path:
-        console.print("[red]npm not found. Please install Node.js >= 18.[/red]")
-        raise typer.Exit(1)
 
     # Find source bridge: first check package data, then source dir
     pkg_bridge = Path(__file__).parent.parent / "bridge"  # nanobot/bridge (installed)
@@ -1304,6 +1296,36 @@ def _get_bridge_dir() -> Path:
         console.print("Try reinstalling: pip install --force-reinstall nanobot")
         raise typer.Exit(1)
 
+    def source_hash(root: Path) -> str:
+        digest = hashlib.sha256()
+        for path in sorted(root.rglob("*")):
+            if not path.is_file():
+                continue
+            rel = path.relative_to(root)
+            if rel.parts and rel.parts[0] in {"node_modules", "dist"}:
+                continue
+            digest.update(rel.as_posix().encode("utf-8"))
+            digest.update(b"\0")
+            digest.update(path.read_bytes())
+            digest.update(b"\0")
+        return digest.hexdigest()
+
+    expected_hash = source_hash(source)
+    current_hash = stamp_file.read_text().strip() if stamp_file.exists() else None
+
+    # Reuse only a bridge built from the currently installed source.
+    if (user_bridge / "dist" / "index.js").exists() and current_hash == expected_hash:
+        return user_bridge
+
+    if (user_bridge / "dist" / "index.js").exists() and current_hash != expected_hash:
+        console.print(f"{__logo__} WhatsApp bridge source changed; rebuilding bridge...")
+
+    # Check for npm
+    npm_path = shutil.which("npm")
+    if not npm_path:
+        console.print("[red]npm not found. Please install Node.js >= 18.[/red]")
+        raise typer.Exit(1)
+
     console.print(f"{__logo__} Setting up bridge...")
 
     # Copy to user directory
@@ -1319,6 +1341,7 @@ def _get_bridge_dir() -> Path:
 
         console.print("  Building...")
         subprocess.run([npm_path, "run", "build"], cwd=user_bridge, check=True, capture_output=True)
+        stamp_file.write_text(expected_hash + "\n")
 
         console.print("[green]✓[/green] Bridge ready\n")
     except subprocess.CalledProcessError as e:

nanobot/providers/anthropic_provider.py

@@ -434,7 +434,7 @@ class AnthropicProvider(LLMProvider):
             )
 
         max_tokens = max(1, max_tokens)
-        thinking_enabled = bool(reasoning_effort)
+        thinking_enabled = bool(reasoning_effort) and reasoning_effort.lower() != "none"
 
         # claude-opus-4-7 deprecated the `temperature` parameter entirely — the
         # API returns 400 if it is present, on any code path.

nanobot/providers/azure_openai_provider.py

@@ -71,7 +71,7 @@ class AzureOpenAIProvider(LLMProvider):
         reasoning_effort: str | None = None,
     ) -> bool:
         """Return True when temperature is likely supported for this deployment."""
-        if reasoning_effort:
+        if reasoning_effort and reasoning_effort.lower() != "none":
             return False
         name = deployment_name.lower()
         return not any(token in name for token in ("gpt-5", "o1", "o3", "o4"))
@@ -102,7 +102,7 @@ class AzureOpenAIProvider(LLMProvider):
         if self._supports_temperature(deployment, reasoning_effort):
             body["temperature"] = temperature
 
-        if reasoning_effort:
+        if reasoning_effort and reasoning_effort.lower() != "none":
             body["reasoning"] = {"effort": reasoning_effort}
             body["include"] = ["reasoning.encrypted_content"]
 

nanobot/providers/openai_codex_provider.py

@@ -60,7 +60,7 @@ class OpenAICodexProvider(LLMProvider):
             "tool_choice": tool_choice or "auto",
             "parallel_tool_calls": True,
         }
-        if reasoning_effort:
+        if reasoning_effort and reasoning_effort.lower() != "none":
             body["reasoning"] = {"effort": reasoning_effort}
         if tools:
             body["tools"] = convert_tools(tools)

nanobot/providers/openai_compat_provider.py

@@ -570,7 +570,7 @@ class OpenAICompatProvider(LLMProvider):
             # DashScope accepts none/minimum/low/medium/high/xhigh; "minimal" 400s.
             wire_effort = "minimum"
 
-        if wire_effort:
+        if wire_effort and semantic_effort != "none":
             kwargs["reasoning_effort"] = wire_effort
 
         # Provider-specific thinking parameters.
@@ -579,7 +579,7 @@ class OpenAICompatProvider(LLMProvider):
         # The mapping is driven by ProviderSpec.thinking_style so that adding
         # a new provider never requires touching this function.
         if spec and spec.thinking_style and reasoning_effort is not None:
-            thinking_enabled = semantic_effort != "minimal"
+            thinking_enabled = semantic_effort not in ("none", "minimal")
             extra = _THINKING_STYLE_MAP.get(spec.thinking_style, lambda _: None)(thinking_enabled)
             if extra:
                 kwargs.setdefault("extra_body", {}).update(extra)
@@ -589,7 +589,7 @@ class OpenAICompatProvider(LLMProvider):
         # so that OpenRouter-style names like "moonshotai/kimi-k2.5" are handled
         # identically to bare names like "kimi-k2.5".
         if reasoning_effort is not None and _is_kimi_thinking_model(model_name):
-            thinking_enabled = semantic_effort != "minimal"
+            thinking_enabled = semantic_effort not in ("none", "minimal")
             kwargs.setdefault("extra_body", {}).update(
                 {"thinking": {"type": "enabled" if thinking_enabled else "disabled"}}
             )
@@ -609,9 +609,9 @@ class OpenAICompatProvider(LLMProvider):
         # thinking happened on that turn").
         thinking_active = (
             (spec and spec.thinking_style and reasoning_effort is not None
-             and semantic_effort != "minimal")
+             and semantic_effort not in ("none", "minimal"))
             or (reasoning_effort is not None and _is_kimi_thinking_model(model_name)
-                and semantic_effort != "minimal")
+                and semantic_effort not in ("none", "minimal"))
         )
         if thinking_active:
             for msg in kwargs["messages"]:

nanobot/providers/registry.py

@@ -267,7 +267,7 @@ PROVIDERS: tuple[ProviderSpec, ...] = (
     # Gemini: Google's OpenAI-compatible endpoint
     ProviderSpec(
         name="gemini",
-        keywords=("gemini",),
+        keywords=("gemini", "gemma"),
         env_key="GEMINI_API_KEY",
         display_name="Gemini",
         backend="openai_compat",

pyproject.toml

@@ -1,12 +1,13 @@
 [project]
 name = "nanobot-ai"
-version = "0.1.5.post2"
+version = "0.1.5.post3"
 description = "A lightweight personal AI assistant framework"
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.11"
 license = {text = "MIT"}
 authors = [
-    {name = "nanobot contributors"}
+    {name = "Xubin Ren"},
+    {name = "the nanobot contributors"}
 ]
 keywords = ["ai", "agent", "chatbot"]
 classifiers = [

tests/agent/test_memory_store.py

@@ -141,6 +141,49 @@ class TestHistoryWithCursor:
         assert len(entries) == 2
         assert entries[0]["cursor"] in {4, 5}
 
+    def test_write_entries_uses_atomic_write(self, tmp_path):
+        """_write_entries uses temp file + os.replace for atomicity."""
+        store = MemoryStore(tmp_path)
+        store.append_history("event 1")
+        store.append_history("event 2")
+        store.append_history("event 3")
+        entries = store.read_unprocessed_history(since_cursor=0)
+
+        # Monitor temp file existence
+        tmp_path_obj = store.history_file.with_suffix(".jsonl.tmp")
+        assert not tmp_path_obj.exists()  # Should not exist initially
+
+        # Call _write_entries
+        store._write_entries(entries)
+
+        # Temp file should be cleaned up
+        assert not tmp_path_obj.exists()
+        # Original file should exist
+        assert store.history_file.exists()
+
+    def test_write_entries_cleans_up_tmp_on_exception(self, tmp_path, monkeypatch):
+        """Exception during _write_entries cleans up the temp file."""
+        store = MemoryStore(tmp_path)
+        store.append_history("event 1")
+        entries = store.read_unprocessed_history(since_cursor=0)
+
+        tmp_path_obj = store.history_file.with_suffix(".jsonl.tmp")
+
+        # Mock os.replace to raise an exception
+        def failing_replace(*args, **kwargs):
+            raise RuntimeError("Simulated failure")
+
+        monkeypatch.setattr('os.replace', failing_replace)
+
+        with pytest.raises(RuntimeError):
+            store._write_entries(entries)
+
+        # Temp file should be cleaned up
+        assert not tmp_path_obj.exists()
+
+        # Original file should still exist (because replace failed)
+        assert store.history_file.exists()
+
 
 class TestAppendHistoryHardCap:
     """append_history has a defensive cap that catches new callers who forgot

tests/agent/tools/test_self_tool_runtime_sync.py

@@ -0,0 +1,29 @@
+"""Focused tests for MyTool runtime sync side effects."""
+
+from unittest.mock import MagicMock
+
+import pytest
+
+from nanobot.agent.tools.self import MyTool
+
+
+@pytest.mark.asyncio
+async def test_my_tool_max_iterations_syncs_subagent_limit() -> None:
+    loop = MagicMock()
+    loop.max_iterations = 40
+    loop._runtime_vars = {}
+    loop.subagents = MagicMock()
+    loop.subagents.max_iterations = loop.max_iterations
+
+    def _sync_subagent_runtime_limits() -> None:
+        loop.subagents.max_iterations = loop.max_iterations
+
+    loop._sync_subagent_runtime_limits = _sync_subagent_runtime_limits
+
+    tool = MyTool(loop=loop)
+
+    result = await tool.execute(action="set", key="max_iterations", value=80)
+
+    assert "Set max_iterations = 80" in result
+    assert loop.max_iterations == 80
+    assert loop.subagents.max_iterations == 80

tests/agent/tools/test_subagent_tools.py

@@ -54,11 +54,129 @@ async def test_subagent_exec_tool_receives_allowed_env_keys(tmp_path):
     mgr.runner.run.assert_awaited_once()
 
 
+@pytest.mark.asyncio
+async def test_subagent_uses_configured_max_iterations(tmp_path):
+    """Subagents should honor the configured tool-iteration limit."""
+    from nanobot.agent.subagent import SubagentManager, SubagentStatus
+    from nanobot.bus.queue import MessageBus
+
+    bus = MessageBus()
+    provider = MagicMock()
+    provider.get_default_model.return_value = "test-model"
+    mgr = SubagentManager(
+        provider=provider,
+        workspace=tmp_path,
+        bus=bus,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+        max_iterations=37,
+    )
+    mgr._announce_result = AsyncMock()
+
+    async def fake_run(spec):
+        assert spec.max_iterations == 37
+        return SimpleNamespace(
+            stop_reason="done",
+            final_content="done",
+            error=None,
+            tool_events=[],
+        )
+
+    mgr.runner.run = AsyncMock(side_effect=fake_run)
+
+    status = SubagentStatus(
+        task_id="sub-1", label="label", task_description="do task", started_at=time.monotonic()
+    )
+    await mgr._run_subagent(
+        "sub-1", "do task", "label", {"channel": "test", "chat_id": "c1"}, status
+    )
+
+    mgr.runner.run.assert_awaited_once()
+
+
+def test_subagent_default_max_iterations_matches_agent_defaults(tmp_path):
+    """Direct SubagentManager construction should use the agent default limit."""
+    from nanobot.agent.subagent import SubagentManager
+    from nanobot.bus.queue import MessageBus
+
+    bus = MessageBus()
+    provider = MagicMock()
+    provider.get_default_model.return_value = "test-model"
+
+    mgr = SubagentManager(
+        provider=provider,
+        workspace=tmp_path,
+        bus=bus,
+        max_tool_result_chars=_MAX_TOOL_RESULT_CHARS,
+    )
+
+    assert mgr.max_iterations == AgentDefaults().max_tool_iterations
+
+
+def test_agent_loop_passes_max_iterations_to_subagents(tmp_path):
+    """AgentLoop's configured limit should be shared with spawned subagents."""
+    from nanobot.agent.loop import AgentLoop
+    from nanobot.bus.queue import MessageBus
+
+    bus = MessageBus()
+    provider = MagicMock()
+    provider.get_default_model.return_value = "test-model"
+
+    loop = AgentLoop(
+        bus=bus,
+        provider=provider,
+        workspace=tmp_path,
+        model="test-model",
+        max_iterations=42,
+    )
+
+    assert loop.subagents.max_iterations == 42
+
+
+@pytest.mark.asyncio
+async def test_agent_loop_syncs_updated_max_iterations_before_run(tmp_path):
+    """Runtime max_iterations changes should be reflected before tool execution."""
+    from nanobot.agent.loop import AgentLoop
+    from nanobot.bus.queue import MessageBus
+
+    bus = MessageBus()
+    provider = MagicMock()
+    provider.get_default_model.return_value = "test-model"
+
+    loop = AgentLoop(
+        bus=bus,
+        provider=provider,
+        workspace=tmp_path,
+        model="test-model",
+        max_iterations=42,
+    )
+    loop.tools.get_definitions = MagicMock(return_value=[])
+
+    async def fake_run(spec):
+        assert spec.max_iterations == 55
+        assert loop.subagents.max_iterations == 55
+        return SimpleNamespace(
+            stop_reason="done",
+            final_content="done",
+            error=None,
+            tool_events=[],
+            messages=[],
+            usage={},
+            had_injections=False,
+            tools_used=[],
+        )
+
+    loop.runner.run = AsyncMock(side_effect=fake_run)
+    loop.max_iterations = 55
+
+    await loop._run_agent_loop([])
+
+    loop.runner.run.assert_awaited_once()
+
+
 @pytest.mark.asyncio
 async def test_drain_pending_blocks_while_subagents_running(tmp_path):
     """_drain_pending should block when no messages are available but sub-agents are still running."""
     from nanobot.agent.loop import AgentLoop
-    from nanobot.agent.subagent import SubagentManager
     from nanobot.bus.events import InboundMessage
     from nanobot.bus.queue import MessageBus
     from nanobot.session.manager import Session
@@ -74,8 +192,6 @@ async def test_drain_pending_blocks_while_subagents_running(tmp_path):
     injection_callback = None
 
     # Capture the injection_callback that _run_agent_loop creates
-    original_run = loop.runner.run
-
     async def fake_runner_run(spec):
         nonlocal injection_callback
         injection_callback = spec.injection_callback

tests/channels/test_channel_manager_delta_coalescing.py

@@ -1,6 +1,6 @@
 """Tests for ChannelManager delta coalescing to reduce streaming latency."""
 import asyncio
-from unittest.mock import AsyncMock, MagicMock
+from unittest.mock import AsyncMock
 
 import pytest
 
@@ -298,6 +298,101 @@ class TestDispatchOutboundWithCoalescing:
         assert pending[0].content == "Final"
 
 
+class TestProgressFiltering:
+    """Progress filtering should honor per-channel settings."""
+
+    def test_progress_visibility_uses_global_defaults(self, manager):
+        assert manager._should_send_progress("mock", tool_hint=False) is True
+        assert manager._should_send_progress("mock", tool_hint=True) is False
+
+    def test_progress_visibility_uses_channel_overrides(self, manager):
+        manager.channels["mock"].send_progress = False
+        manager.channels["mock"].send_tool_hints = True
+
+        assert manager._should_send_progress("mock", tool_hint=False) is False
+        assert manager._should_send_progress("mock", tool_hint=True) is True
+
+    def test_progress_visibility_returns_false_for_missing_channel(self, manager):
+        assert manager._should_send_progress("nonexistent", tool_hint=False) is False
+        assert manager._should_send_progress("nonexistent", tool_hint=True) is False
+
+    def test_resolve_bool_override_dict(self, manager):
+        assert manager._resolve_bool_override({}, "send_progress", True) is True
+        assert manager._resolve_bool_override({"send_progress": False}, "send_progress", True) is False
+        assert manager._resolve_bool_override({"sendProgress": False}, "send_progress", True) is False
+        assert manager._resolve_bool_override({"send_progress": "false"}, "send_progress", True) is True
+
+    def test_resolve_bool_override_model(self, manager):
+        class FakeSection:
+            send_progress = False
+            send_tool_hints = True
+
+        assert manager._resolve_bool_override(FakeSection(), "send_progress", True) is False
+        assert manager._resolve_bool_override(FakeSection(), "send_tool_hints", False) is True
+        # Missing attribute falls back to default
+        assert manager._resolve_bool_override(FakeSection(), "unknown_key", True) is True
+
+    @pytest.mark.asyncio
+    async def test_channel_override_can_drop_progress_message(self, manager, bus):
+        manager.channels["mock"].send_progress = False
+        await bus.publish_outbound(OutboundMessage(
+            channel="mock",
+            chat_id="chat1",
+            content="thinking",
+            metadata={"_progress": True},
+        ))
+        await bus.publish_outbound(OutboundMessage(
+            channel="mock",
+            chat_id="chat1",
+            content="final answer",
+            metadata={},
+        ))
+
+        task = asyncio.create_task(manager._dispatch_outbound())
+        try:
+            for _ in range(30):
+                if manager.channels["mock"]._send_mock.await_count >= 1:
+                    break
+                await asyncio.sleep(0.05)
+        finally:
+            task.cancel()
+            try:
+                await task
+            except asyncio.CancelledError:
+                pass
+
+        send_mock = manager.channels["mock"]._send_mock
+        assert send_mock.await_count == 1
+        assert send_mock.await_args_list[0].args[0].content == "final answer"
+
+    @pytest.mark.asyncio
+    async def test_channel_override_can_enable_tool_hints(self, manager, bus):
+        manager.channels["mock"].send_tool_hints = True
+        await bus.publish_outbound(OutboundMessage(
+            channel="mock",
+            chat_id="chat1",
+            content="read_file(foo.py)",
+            metadata={"_progress": True, "_tool_hint": True},
+        ))
+
+        task = asyncio.create_task(manager._dispatch_outbound())
+        try:
+            for _ in range(30):
+                if manager.channels["mock"]._send_mock.await_count >= 1:
+                    break
+                await asyncio.sleep(0.05)
+        finally:
+            task.cancel()
+            try:
+                await task
+            except asyncio.CancelledError:
+                pass
+
+        send_mock = manager.channels["mock"]._send_mock
+        assert send_mock.await_count == 1
+        assert send_mock.await_args_list[0].args[0].content == "read_file(foo.py)"
+
+
 class TestRetryWaitFiltering:
     """Internal provider retry heartbeats must never reach channels."""
 

tests/channels/test_feishu_reply.py

@@ -580,6 +580,32 @@ async def test_reply_without_reply_in_thread_when_disabled() -> None:
     channel._client.im.v1.message.create.assert_called_once()
 
 
+@pytest.mark.asyncio
+async def test_topic_reply_does_not_force_reply_in_thread_when_disabled() -> None:
+    """Topic replies must not create new Feishu topics when reply_to_message is False."""
+    channel = _make_feishu_channel(reply_to_message=False)
+
+    reply_resp = MagicMock()
+    reply_resp.success.return_value = True
+    channel._client.im.v1.message.reply.return_value = reply_resp
+
+    await channel.send(OutboundMessage(
+        channel="feishu",
+        chat_id="oc_abc",
+        content="hello",
+        metadata={
+            "message_id": "om_child456",
+            "chat_type": "group",
+            "thread_id": "om_root123",
+        },
+    ))
+
+    channel._client.im.v1.message.reply.assert_called_once()
+    call_args = channel._client.im.v1.message.reply.call_args
+    request = call_args[0][0]
+    assert request.request_body.reply_in_thread is not True
+
+
 @pytest.mark.asyncio
 async def test_reply_keeps_fallback_when_reply_fails() -> None:
     """Even with reply_to_message=True, fallback to create on reply failure."""

tests/channels/test_feishu_streaming.py

@@ -10,13 +10,14 @@ from nanobot.bus.queue import MessageBus
 from nanobot.channels.feishu import FeishuChannel, FeishuConfig, _FeishuStreamBuf
 
 
-def _make_channel(streaming: bool = True) -> FeishuChannel:
+def _make_channel(streaming: bool = True, reply_to_message: bool = False) -> FeishuChannel:
     config = FeishuConfig(
         enabled=True,
         app_id="cli_test",
         app_secret="secret",
         allow_from=["*"],
         streaming=streaming,
+        reply_to_message=reply_to_message,
     )
     ch = FeishuChannel(config, MessageBus())
     ch._client = MagicMock()
@@ -148,6 +149,62 @@ class TestSendDelta:
         ch._client.im.v1.message.create.assert_called_once()
         ch._client.cardkit.v1.card_element.content.assert_called_once()
 
+    @pytest.mark.asyncio
+    async def test_group_delta_uses_create_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        ch._client.cardkit.v1.card.create.return_value = _mock_create_card_response("card_new")
+        ch._client.im.v1.message.create.return_value = _mock_send_response("om_new")
+        ch._client.cardkit.v1.card_element.content.return_value = _mock_content_response()
+
+        await ch.send_delta(
+            "oc_chat1",
+            "Hello ",
+            metadata={"message_id": "om_001", "chat_type": "group"},
+        )
+
+        ch._client.im.v1.message.create.assert_called_once()
+        ch._client.im.v1.message.reply.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_group_delta_keeps_existing_topic_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        ch._client.cardkit.v1.card.create.return_value = _mock_create_card_response("card_new")
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+        ch._client.cardkit.v1.card_element.content.return_value = _mock_content_response()
+
+        await ch.send_delta(
+            "oc_chat1",
+            "Hello ",
+            metadata={"message_id": "om_001", "chat_type": "group", "thread_id": "ot_001"},
+        )
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is not True
+
+    @pytest.mark.asyncio
+    async def test_group_delta_replies_in_thread_when_reply_enabled(self):
+        ch = _make_channel(reply_to_message=True)
+        ch._client.cardkit.v1.card.create.return_value = _mock_create_card_response("card_new")
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+        ch._client.cardkit.v1.card_element.content.return_value = _mock_content_response()
+
+        await ch.send_delta(
+            "oc_chat1",
+            "Hello ",
+            metadata={"message_id": "om_001", "chat_type": "group"},
+        )
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is True
+
     @pytest.mark.asyncio
     async def test_second_delta_within_interval_skips_update(self):
         ch = _make_channel()
@@ -204,6 +261,70 @@ class TestSendDelta:
         ch._client.cardkit.v1.card_element.content.assert_not_called()
         ch._client.im.v1.message.create.assert_called_once()
 
+    @pytest.mark.asyncio
+    async def test_stream_end_fallback_group_uses_create_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        ch._stream_bufs["om_001"] = _FeishuStreamBuf(
+            text="Fallback content", card_id=None, sequence=0, last_edit=0.0,
+        )
+        ch._client.im.v1.message.create.return_value = _mock_send_response("om_fb")
+
+        await ch.send_delta(
+            "oc_chat1",
+            "",
+            metadata={"_stream_end": True, "message_id": "om_001", "chat_type": "group"},
+        )
+
+        ch._client.im.v1.message.create.assert_called_once()
+        ch._client.im.v1.message.reply.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_stream_end_fallback_keeps_existing_topic_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        ch._stream_bufs["om_001"] = _FeishuStreamBuf(
+            text="Fallback content", card_id=None, sequence=0, last_edit=0.0,
+        )
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+
+        await ch.send_delta(
+            "oc_chat1",
+            "",
+            metadata={
+                "_stream_end": True,
+                "message_id": "om_001",
+                "chat_type": "group",
+                "thread_id": "ot_001",
+            },
+        )
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is not True
+
+    @pytest.mark.asyncio
+    async def test_stream_end_fallback_group_replies_when_reply_enabled(self):
+        ch = _make_channel(reply_to_message=True)
+        ch._stream_bufs["om_001"] = _FeishuStreamBuf(
+            text="Fallback content", card_id=None, sequence=0, last_edit=0.0,
+        )
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+
+        await ch.send_delta(
+            "oc_chat1",
+            "",
+            metadata={"_stream_end": True, "message_id": "om_001", "chat_type": "group"},
+        )
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is True
+
     @pytest.mark.asyncio
     async def test_stream_end_fallback_when_final_update_fails(self):
         """If streaming mode was closed (e.g. Feishu timeout), fall back to a regular card."""
@@ -316,6 +437,64 @@ class TestToolHintInlineStreaming:
         assert "oc_chat1" not in ch._stream_bufs
         ch._client.im.v1.message.create.assert_called_once()
 
+    @pytest.mark.asyncio
+    async def test_tool_hint_group_uses_create_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        ch._client.im.v1.message.create.return_value = _mock_send_response("om_hint")
+
+        msg = OutboundMessage(
+            channel="feishu", chat_id="oc_chat1",
+            content='read_file("path")',
+            metadata={"_tool_hint": True, "message_id": "om_001", "chat_type": "group"},
+        )
+        await ch.send(msg)
+
+        ch._client.im.v1.message.create.assert_called_once()
+        ch._client.im.v1.message.reply.assert_not_called()
+
+    @pytest.mark.asyncio
+    async def test_tool_hint_keeps_existing_topic_when_reply_disabled(self):
+        ch = _make_channel(reply_to_message=False)
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+
+        msg = OutboundMessage(
+            channel="feishu", chat_id="oc_chat1",
+            content='read_file("path")',
+            metadata={
+                "_tool_hint": True,
+                "message_id": "om_001",
+                "chat_type": "group",
+                "thread_id": "ot_001",
+            },
+        )
+        await ch.send(msg)
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is not True
+
+    @pytest.mark.asyncio
+    async def test_tool_hint_group_replies_when_reply_enabled(self):
+        ch = _make_channel(reply_to_message=True)
+        reply_resp = MagicMock()
+        reply_resp.success.return_value = True
+        ch._client.im.v1.message.reply.return_value = reply_resp
+
+        msg = OutboundMessage(
+            channel="feishu", chat_id="oc_chat1",
+            content='read_file("path")',
+            metadata={"_tool_hint": True, "message_id": "om_001", "chat_type": "group"},
+        )
+        await ch.send(msg)
+
+        ch._client.im.v1.message.reply.assert_called_once()
+        ch._client.im.v1.message.create.assert_not_called()
+        request = ch._client.im.v1.message.reply.call_args[0][0]
+        assert request.request_body.reply_in_thread is True
+
     @pytest.mark.asyncio
     async def test_consecutive_tool_hints_append(self):
         """When multiple tool hints arrive consecutively, each appends to the card."""

tests/providers/test_anthropic_thinking.py

@@ -83,3 +83,10 @@ def test_opus_4_7_omits_temperature_none() -> None:
     kw = _build(_make_provider("claude-opus-4-7"), None)
     assert "temperature" not in kw
     assert "thinking" not in kw
+
+
+def test_reasoning_effort_string_none_does_not_enable_thinking() -> None:
+    """reasoning_effort='none' must not enable thinking — treated same as disabled."""
+    kw = _build(_make_provider(), "none")
+    assert "thinking" not in kw
+    assert kw["temperature"] == 0.7

tests/providers/test_azure_openai_provider.py

@@ -78,6 +78,11 @@ def test_supports_temperature_with_reasoning_effort():
     assert AzureOpenAIProvider._supports_temperature("gpt-4o", reasoning_effort="medium") is False
 
 
+def test_supports_temperature_with_reasoning_effort_none_string():
+    """reasoning_effort='none' must NOT suppress temperature — it means thinking is off."""
+    assert AzureOpenAIProvider._supports_temperature("gpt-4o", reasoning_effort="none") is True
+
+
 # ---------------------------------------------------------------------------
 # _build_body — Responses API body construction
 # ---------------------------------------------------------------------------
@@ -131,6 +136,16 @@ def test_build_body_with_reasoning():
     assert "temperature" not in body
 
 
+def test_build_body_reasoning_effort_none_string_omits_reasoning():
+    """reasoning_effort='none' must not inject a reasoning body and must allow temperature."""
+    provider = AzureOpenAIProvider(api_key="k", api_base="https://r.com", default_model="gpt-4o")
+    body = provider._build_body(
+        [{"role": "user", "content": "hi"}], None, "gpt-4o", 4096, 0.7, "none", None,
+    )
+    assert "reasoning" not in body
+    assert body["temperature"] == 0.7
+
+
 def test_build_body_image_conversion():
     """image_url content blocks should be converted to input_image."""
     provider = AzureOpenAIProvider(api_key="k", api_base="https://r.com", default_model="gpt-4o")

tests/providers/test_litellm_kwargs.py

@@ -121,6 +121,14 @@ def test_openrouter_spec_is_gateway() -> None:
     assert spec.default_api_base == "https://openrouter.ai/api/v1"
 
 
+def test_gemma_routes_to_gemini_provider() -> None:
+    """gemma models (e.g. gemma-3-27b-it) must auto-route to Gemini when GEMINI_API_KEY is set.
+    Users running gemma via the Gemini API endpoint expect automatic provider detection."""
+    spec = find_by_name("gemini")
+    assert spec is not None
+    assert "gemma" in spec.keywords
+
+
 def test_openrouter_sets_default_attribution_headers() -> None:
     spec = find_by_name("openrouter")
     with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI") as MockClient:
@@ -1050,3 +1058,46 @@ def test_kimi_k2_thinking_series_no_thinking_injection() -> None:
     """kimi-k2-thinking series models must NOT receive extra_body.thinking."""
     kw = _build_kwargs_for("moonshot", "kimi-k2-thinking", reasoning_effort="high")
     assert "extra_body" not in kw
+
+
+# ---------------------------------------------------------------------------
+# reasoning_effort="none" — treated as thinking disabled 
+# ---------------------------------------------------------------------------
+
+def test_deepseek_thinking_disabled_for_none_string() -> None:
+    """reasoning_effort='none' must send thinking.type=disabled and skip reasoning_effort field."""
+    kw = _build_kwargs_for("deepseek", "deepseek-v4-pro", reasoning_effort="none")
+    assert kw.get("extra_body") == {"thinking": {"type": "disabled"}}
+    assert "reasoning_effort" not in kw
+
+
+def test_kimi_k25_thinking_disabled_for_none_string() -> None:
+    """reasoning_effort='none' maps to thinking disabled for kimi-k2.5."""
+    kw = _build_kwargs_for("moonshot", "kimi-k2.5", reasoning_effort="none")
+    assert kw.get("extra_body") == {"thinking": {"type": "disabled"}}
+
+
+def test_dashscope_thinking_disabled_for_none_string() -> None:
+    """reasoning_effort='none' disables thinking and must not emit reasoning_effort on DashScope."""
+    kw = _build_kwargs_for("dashscope", "qwen3.6-plus", reasoning_effort="none")
+    assert kw.get("extra_body") == {"enable_thinking": False}
+    assert "reasoning_effort" not in kw
+
+
+def test_deepseek_no_backfill_when_reasoning_effort_none_string() -> None:
+    """reasoning_effort='none' must NOT trigger reasoning_content backfill (thinking inactive)."""
+    spec = find_by_name("deepseek")
+    with patch("nanobot.providers.openai_compat_provider.AsyncOpenAI"):
+        p = OpenAICompatProvider(api_key="k", default_model="deepseek-v4-pro", spec=spec)
+    messages = [
+        {"role": "user", "content": "hi"},
+        {"role": "assistant", "content": "ok"},
+        {"role": "user", "content": "continue"},
+    ]
+    kw = p._build_kwargs(
+        messages=list(messages), tools=None, model="deepseek-v4-pro",
+        max_tokens=1024, temperature=0.7,
+        reasoning_effort="none", tool_choice=None,
+    )
+    assistant = kw["messages"][1]
+    assert "reasoning_content" not in assistant