Root cause: when consolidation LLM fails, raw_archive() dumped full message
content (~1MB) into history.jsonl with no size limit. Since build_system_prompt()
injects history.jsonl into every system prompt, all subsequent LLM calls exceeded
the 200K context window with error 1261.
Additionally, _cap_consolidation_boundary's 60-message cap caused consolidation
to get stuck on sessions with long tool chains (200+ iterations), triggering
the raw_archive fallback in the first place.
Three-layer fix:
- Remove _cap_consolidation_boundary: let pick_consolidation_boundary drive
chunk sizing based solely on token budget
- Truncate archive() input: use tiktoken to cap formatted text to the model's
input token budget before sending to consolidation LLM
- Truncate raw_archive() output: cap history.jsonl entries at 16K chars
When the main agent spawns multiple sub-agents, each completion
independently triggered a new _dispatch, causing 3-4 user-visible
responses instead of a single comprehensive report.
- Extend _drain_pending to block-wait on pending_queue when sub-agents
are still running, keeping the runner loop alive for in-order injection
- Pass pending_queue in the system message path so subsequent sub-agent
results can still be injected mid-turn via a new dispatch
Move the non-int cursor guard out of the two consumer sites and into a
shared ``_iter_valid_entries`` iterator so the invariant lives in one
place. Closes three gaps left by the original fix:
* ``bool`` is now rejected — ``isinstance(True, int)`` is ``True`` in
Python, so the previous guard silently treated ``{"cursor": true}`` as
cursor ``1``.
* Recovery now returns ``max(valid cursors) + 1``. Under adversarial
corruption "first int scanning in reverse" is not the same thing, and
only ``max`` keeps the recovered cursor strictly greater than every
legitimate cursor still on disk.
* Non-int cursors are logged exactly once per ``MemoryStore``. Silently
dropping corrupted entries hides the root cause (an external writer
to ``memory/history.jsonl``); rate-limiting keeps the log clean when
the same poisoned file is read every turn.
All 7 tests from the original fix pass unchanged; 3 new tests pin the
invariants above.
Made-with: Cursor
_next_cursor now checks isinstance(cursor, int) before arithmetic,
falling back to a reverse scan of all entries when the last entry's
cursor is corrupted. read_unprocessed_history skips entries with
non-int cursors instead of crashing on comparison.
Root cause: external callers (cron jobs, plugins) occasionally wrote
string cursors to history.jsonl, which blocked all subsequent
append_history calls with TypeError/ValueError.
Includes 7 regression tests covering string, float, null, and list
cursor types.
The retry branch is only reachable via `except Exception`, and
`CancelledError` inherits from `BaseException`, so today it naturally
bypasses the retry path and /stop still works. Add one focused
regression test so any future refactor that widens the retry catch to
`BaseException`, re-orders the handlers, or adds `CancelledError` to
`_TRANSIENT_EXC_NAMES` fails CI instead of silently swallowing /stop.
Made-with: Cursor
When an MCP server restarts or a network connection drops between
tool calls, the existing session throws ClosedResourceError,
BrokenPipeError, ConnectionResetError, etc. Currently these are
caught as generic exceptions and returned as permanent failures
to the LLM, which then tells the user 'my tools are broken.'
This change adds a single automatic retry with a 1-second backoff
for transient connection-class errors in MCPToolWrapper,
MCPResourceWrapper, and MCPPromptWrapper. Non-transient errors
(ValueError, RuntimeError, McpError, etc.) are not retried.
The retry is conservative:
- Only 1 retry (not configurable, to keep the change minimal)
- Only for a specific set of connection-class exceptions
- Matched by exception class name to avoid importing anyio/etc.
- 1s sleep between attempts to allow the server to recover
- Clear logging distinguishes retried vs permanent failures
In production this eliminates most 'MCP tool call failed:
ClosedResourceError' noise when MCP bridge processes restart
(e.g. after config changes or OOM kills).
Tests: 22 new tests covering retry, exhaustion, non-transient
bypass, timeout bypass, and all three wrapper types.
Add a regression test that actually runs the CancelledError branch of
AgentLoop._dispatch end-to-end and asserts the in-flight checkpoint is
materialized into session.messages before the cancellation unwinds.
The three existing tests call _restore_runtime_checkpoint directly, so
they pass even if the cancel-time restore is ever removed from
_dispatch. This new test is the one that actually locks the fix in
place.
Made-with: Cursor
When a user sends /stop to interrupt an active agent turn, the task is
cancelled via CancelledError. Previously, the cancellation handler just
logged and re-raised, discarding any tool results and assistant messages
accumulated during the interrupted turn.
The runtime checkpoint mechanism already persists partial turn state
(assistant messages, completed tool results, pending tool calls) into
session metadata via _emit_checkpoint. However, this checkpoint was only
materialized into session history on the NEXT incoming message via
_restore_runtime_checkpoint — not at cancellation time.
Now the CancelledError handler in _dispatch calls
_restore_runtime_checkpoint immediately, so the partial context is
preserved in session history. This means the next message the user sends
will see all the work that was done before /stop, rather than starting
from scratch.
Fixes#2966
Includes 3 tests verifying checkpoint restoration on cancellation.
`append_history` previously used `strip_think(entry) or entry.rstrip()`
as a safety net, so if the entire entry was a template-token leak (e.g.
`<think>reasoning</think>` or `<channel|>` alone), the raw leaked text
was still persisted to history — later re-introducing the very content
`strip_think` was meant to scrub, via consolidation / replay.
Persist the cleaned content directly. When cleanup empties a non-empty
entry, log at debug and store an empty-content record (cursor continuity
preserved). Adds 3 regression tests in test_memory_store.py covering:
- Well-formed thinking blocks are stripped before persistence.
- Pure-leak entries persist as empty, not as raw text.
- Malformed prefix leaks (`<channel|>`) also persist as empty.
The previous fix hardcoded session_key_override as channel:chat_id which
broke unified session mode where pending queues use "unified:default".
Propagate the effective key from _set_tool_context through SpawnTool
into the origin dict so _announce_result routes to the correct pending
queue in both normal and unified session modes.
SessionManager.save() previously used bare open("w") which could
truncate the JSONL file if the process crashed mid-write. Now writes
to a .tmp file and atomically replaces via os.replace(), matching the
pattern already used in qq.py.
_load() now attempts _repair() before returning None, recovering
valid lines from partially-written files. 12 new tests cover atomic
save correctness, temp-file cleanup on failure, and repair of
truncated/corrupt JSONL.
cowork-with:opencode(glm-5.1)
Literal["standard", "persistent"] fields are now rendered as select
dropdowns instead of free-text input. This makes provider_retry_mode
and any future Literal fields self-documenting in the wizard.
- Add [H] Channel Common menu to configure send_progress, send_tool_hints,
send_max_retries, and transcription_provider
- Add [I] API Server menu to configure host, port, timeout
- Add real-time Pydantic field constraint validation (ge/gt/le/lt/min_length/max_length)
with constraint hints shown in field display (e.g. "Send Max Retries (0-10)")
- Add _pause() to View Configuration Summary to prevent immediate screen clear
- Fix _format_value dict branch to handle BaseModel instances without crashing
Move all behavioral instructions out of identity.md into SOUL.md so that
each file has a single clear purpose:
- identity.md: capability facts only (runtime, workspace, format hints,
tool guidance, untrusted content warning)
- SOUL.md: behavioral rules (name, personality, execution rules)
The "Act, don't narrate" rule is refined into layered behavior: act
immediately on single-step tasks, plan first for multi-step tasks. This
eliminates the contradiction where identity said "never end with a plan"
but user SOUL.md said "always plan first".
Add two focused regression tests for the retry-wait leak this PR fixes:
- tests/agent/test_runner.py::test_runner_binds_on_retry_wait_to_retry_callback_not_progress
locks in that `AgentRunSpec.retry_wait_callback` (not `progress_callback`) is
what `_build_request_kwargs` forwards to the provider as `on_retry_wait`.
- tests/channels/test_channel_manager_delta_coalescing.py::TestRetryWaitFiltering
runs `_dispatch_outbound` end-to-end and asserts that `_retry_wait: True`
messages never reach channel send.
Both tests fail on origin/main and pass with this PR's fix applied.
Made-with: Cursor
- Add inline rationale for persisting before ContextBuilder and for
passing current_message="" on subagent follow-ups (avoids
double-projection after merge).
- Skip persistence for empty subagent content (no-op messages should
not pollute history).
- Add regression test covering the empty-content guard.
Made-with: Cursor
MyTool blocks direct access to sensitive nested paths, but its formatter
still printed scalar fields for small config objects. That let
`my(action="check", key="web_config.search")` expose `api_key` in plain
text even though the docs promise sensitive sub-fields are protected.
This keeps the change narrow: sensitive nested config fields are omitted
from MyTool's formatted output, and regression coverage locks the
behavior in.
Constraint: Must preserve existing read-only inspection behavior for non-sensitive fields
Constraint: Keep scope limited to MyTool rather than introducing broader redaction plumbing
Rejected: Rework global context/tool redaction around MyTool | broader than needed for the leak path
Confidence: high
Scope-risk: narrow
Reversibility: clean
Directive: If more nested config rendering is added later, filter sensitive field names at the formatter boundary as well as the path resolver
Tested: PYTHONPATH=$PWD pytest -q tests/agent/tools/test_self_tool.py /Users/jh0927/Workspace/nanobot-validation-artifacts-2026-04-18/test_my_tool_secret_leak_regression.py
Not-tested: Full repository test suite
Related: #3259
When chat_with_retry returns an error response (finish_reason='error')
instead of raising an exception, archive() previously treated the error
message as a valid summary and wrote it to history.jsonl, while the
original session data was already cleared by /new — causing irreversible
data loss.
Fix: check finish_reason after the LLM call and raise RuntimeError on
error responses, which naturally falls through to the existing raw_archive
fallback. This preserves the original messages in history.jsonl instead
of losing them.
Fixes#3244
- Extract synthetic user message string to module-level constant
- Tighten comments in _snip_history recovery branch
- Strengthen no-user edge case test to verify safety net interaction
When _snip_history truncates the message history and the only user message
ends up outside the kept window, providers like GLM reject the resulting
system→assistant sequence with error 1214 ("messages 参数非法").
Two-layer fix:
1. _snip_history now walks backwards through non_system messages to recover
the nearest user message when none exists in the kept window.
2. _enforce_role_alternation inserts a synthetic user message
"(conversation continued)" when the first non-system message is a bare
assistant (no tool_calls), serving as a safety net for any edge cases
that slip through.
Co-authored-by: darlingbud <darlingbud@users.noreply.github.com>
Follow-up to #3212, fully backward compatible:
- Extract the 14-day staleness threshold as `_STALE_THRESHOLD_DAYS` module
constant and pass it into the Phase 1 prompt template as
`{{ stale_threshold_days }}`. The number lived in three places before
(code threshold, prompt instruction, docstring); now there is one.
- Add `DreamConfig.annotate_line_ages` (default True = current behavior)
and propagate it through `Dream.__init__` and the gateway wiring in
cli/commands.py. Gives users a knob to disable the feature without a
code patch if an LLM reacts poorly to the `← Nd` suffix.
- Harden `_annotate_with_ages` against dirty working trees: when HEAD
blob line count disagrees with the working-tree content length, skip
annotation entirely instead of assigning ages to the wrong lines. The
previous `i >= len(ages)` guard only handled one direction of the
mismatch.
- Inline-comment the `max_iterations` 10→15 bump with a pointer to
exp002 so future blame has context.
- Add 4 regression tests: end-to-end `← 30d` reaches prompt, 14/15
threshold boundary, `annotate_line_ages=False` bypasses git entirely
(verified via `assert_not_called`), length-mismatch defense, and
template-var rendering.
Made-with: Cursor
Three improvements to Dream's memory consolidation:
1. Per-line git-blame age annotations: MEMORY.md lines get `← Nd` suffixes
(N>14) from dulwich annotate. SOUL.md/USER.md excluded as permanent.
LLM uses content judgment, not just age, to decide what to prune.
2. Dedup-aware Phase 1 prompt: reframed as dual-task (extract facts +
deduplicate existing files) with explicit redundancy patterns to scan for.
Validated through 20 experiments (exp-002 prompt + max_iter=15 was best,
averaging -1643 chars/5.4% compression per run).
3. Phase 1 analysis as commit body: dream git commits now include the full
Phase 1 analysis for transparency via /dream-log.
4. max_iterations raised from 10 to 15: 30% improvement over 10 with no
risk; 20 showed diminishing returns (exp-020: -701 vs exp-017: -1643).
Add a built-in tool that lets the agent inspect and modify its own
runtime state (model, iterations, context window, etc.).
Key features:
- inspect: view current config, usage stats, and subagent status
- modify: adjust parameters at runtime (protected by type/range validation)
- Subagent observability: inspect running subagent tasks (phase,
iteration, tool events, errors) — subagents are no longer a black box
- Watchdog corrects out-of-bounds values on each iteration
- Enabled by default in read-only mode (self_modify: false)
- All changes are in-memory only; restart restores defaults
- Comprehensive test suite (90 tests)
Includes a self-awareness skill (always-on) with progressive disclosure:
SKILL.md for core rules, references/examples.md for detailed scenarios.
- Convert skills summary from verbose XML (4-5 lines/skill) to compact
markdown list (1 line/skill) with inline path for read_file lookup
- Exclude always-loaded skills (e.g. memory) from the skills index to
avoid duplicating content already in the Active Skills section
- Skip injecting the Memory section when MEMORY.md still matches the
bundled template (i.e. Dream hasn't populated it yet)
The hand-rolled line-by-line YAML parser treated each line independently,
so YAML multiline scalars (folded `>` and literal `|`) were captured as
the literal characters ">" or "|" instead of the actual text content.
Keep late follow-up injections observable when they are drained during max-iteration shutdown so loop-level response suppression still makes the right decision.
Made-with: Cursor
- Migrate "after tools" inline drain to use _try_drain_injections,
completing the refactoring (all 6 drain sites now use the helper).
- Move checkpoint emission into _try_drain_injections via optional
iteration parameter, eliminating the leaky split between helper
and caller for the final-response path.
- Extract _make_injection_callback() test helper to replace 7
identical inject_cb function bodies.
- Add test_injection_cycle_cap_on_error_path to verify the cycle
cap is enforced on error exit paths.
When the agent runner exits due to LLM error, tool error, empty response,
or max_iterations, it breaks out of the iteration loop without draining
the pending injection queue. This causes leftover messages to be
re-published as independent inbound messages, resulting in duplicate or
confusing replies to the user.
Extract the injection drain logic into a `_try_drain_injections` helper
and call it before each break in the error/edge-case paths. If injections
are found, continue the loop instead of breaking. For max_iterations
(where the loop is exhausted), drain injections to prevent re-publish
without continuing.
Prevent proactive compaction from archiving sessions that have an
in-flight agent task, avoiding mid-turn context truncation when a
task runs longer than the idle TTL.
Track text-only user messages that were flushed before the turn loop completes, then materialize an interrupted assistant placeholder on the next request so session history stays legal and later turns do not skip their own assistant reply.
Made-with: Cursor
Use session.add_message for the pre-turn user-message flush and add focused regression tests for crash-time persistence and duplicate-free successful saves.
Made-with: Cursor
Point Dream skill creation at a readable builtin skill-creator template, keep skill writes rooted at the workspace, and document the new skill discovery behavior in README.
Made-with: Cursor
