Allow setting provider params (e.g. cacheRetention) once at the
agents.defaults level instead of repeating them per-model. The merge
order is now: defaults.params -> defaults.models[key].params ->
list[agentId].params.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Problem: When LLM stops responding, the agent hangs for ~5 minutes with no feedback.
Users had to use /stop to recover.
Solution: Add idle timeout detection for LLM streaming responses.
* fix: make cleanup "keep" persist subagent sessions indefinitely
* feat: expose subagent session metadata in sessions list
* fix: include status and timing in sessions_list tool
* fix: hide injected timestamp prefixes in chat ui
* feat: push session list updates over websocket
* feat: expose child subagent sessions in subagents list
* feat: add admin http endpoint to kill sessions
* Emit session.message websocket events for transcript updates
* Estimate session costs in sessions list
* Add direct session history HTTP and SSE endpoints
* Harden dashboard session events and history APIs
* Add session lifecycle gateway methods
* Add dashboard session API improvements
* Add dashboard session model and parent linkage support
* fix: tighten dashboard session API metadata
* Fix dashboard session cost metadata
* Persist accumulated session cost
* fix: stop followup queue drain cfg crash
* Fix dashboard session create and model metadata
* fix: stop guessing session model costs
* Gateway: cache OpenRouter pricing for configured models
* Gateway: add timeout session status
* Fix subagent spawn test config loading
* Gateway: preserve operator scopes without device identity
* Emit user message transcript events and deduplicate plugin warnings
* feat: emit sessions.changed lifecycle event on subagent spawn
Adds a session-lifecycle-events module (similar to transcript-events)
that emits create events when subagents are spawned. The gateway
server.impl.ts listens for these events and broadcasts sessions.changed
with reason=create to SSE subscribers, so dashboards can pick up new
subagent sessions without polling.
* Gateway: allow persistent dashboard orchestrator sessions
* fix: preserve operator scopes for token-authenticated backend clients
Backend clients (like agent-dashboard) that authenticate with a valid gateway
token but don't present a device identity were getting their scopes stripped.
The scope-clearing logic ran before checking the device identity decision,
so even when evaluateMissingDeviceIdentity returned 'allow' (because
roleCanSkipDeviceIdentity passed for token-authed operators), scopes were
already cleared.
Fix: also check decision.kind before clearing scopes, so token-authenticated
operators keep their requested scopes.
* Gateway: allow operator-token session kills
* Fix stale active subagent status after follow-up runs
* Fix dashboard image attachments in sessions send
* Fix completed session follow-up status updates
* feat: stream session tool events to operator UIs
* Add sessions.steer gateway coverage
* Persist subagent timing in session store
* Fix subagent session transcript event keys
* Fix active subagent session status in gateway
* bump session label max to 512
* Fix gateway send session reactivation
* fix: publish terminal session lifecycle state
* feat: change default session reset to effectively never
- Change DEFAULT_RESET_MODE from "daily" to "idle"
- Change DEFAULT_IDLE_MINUTES from 60 to 0 (0 = disabled/never)
- Allow idleMinutes=0 through normalization (don't clamp to 1)
- Treat idleMinutes=0 as "no idle expiry" in evaluateSessionFreshness
- Default behavior: mode "idle" + idleMinutes 0 = sessions never auto-reset
- Update test assertion for new default mode
* fix: prep session management followups (#50101) (thanks @clay-datacurve)
---------
Co-authored-by: Tyler Yust <TYTYYUST@YAHOO.COM>
* feat: make compaction timeout configurable via agents.defaults.compaction.timeoutSeconds
The hardcoded 5-minute (300s) compaction timeout causes large sessions
to enter a death spiral where compaction repeatedly fails and the
session grows indefinitely. This adds agents.defaults.compaction.timeoutSeconds
to allow operators to override the compaction safety timeout.
Default raised to 900s (15min) which is sufficient for sessions up to
~400k tokens. The resolved timeout is also used for the session write
lock duration so locks don't expire before compaction completes.
Fixes#38233
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* test: add resolveCompactionTimeoutMs tests
Cover config resolution edge cases: undefined config, missing
compaction section, valid seconds, fractional values, zero,
negative, NaN, and Infinity.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: add timeoutSeconds to compaction Zod schema
The compaction object schema uses .strict(), so setting the new
timeoutSeconds config option would fail validation at startup.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: enforce integer constraint on compaction timeoutSeconds schema
Prevents sub-second values like 0.5 which would floor to 0ms and
cause immediate compaction timeout. Matches pattern of other
integer timeout fields in the schema.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: clamp compaction timeout to Node timer-safe maximum
Values above ~2.1B ms overflow Node's setTimeout to 1ms, causing
immediate timeout. Clamp to MAX_SAFE_TIMEOUT_MS matching the
pattern in agents/timeout.ts.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: add FIELD_LABELS entry for compaction timeoutSeconds
Maintains label/help parity invariant enforced by
schema.help.quality.test.ts.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: align compaction timeouts with abort handling
* fix: land compaction timeout handling (#46889) (thanks @asyncjason)
---------
Co-authored-by: Jason Separovic <jason@wilma.dog>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Ayaan Zaidi <hi@obviy.us>
Reuses the cron isolated session pattern (resolveCronSession with forceNew)
to give each heartbeat a fresh session with no prior conversation history.
Reduces per-heartbeat token cost from ~100K to ~2-5K tokens.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* feat: add PDF analysis tool with native provider support
New `pdf` tool for analyzing PDF documents with model-powered analysis.
Architecture:
- Native PDF path: sends raw PDF bytes directly to providers that support
inline document input (Anthropic via DocumentBlockParam, Google Gemini
via inlineData with application/pdf MIME type)
- Extraction fallback: for providers without native PDF support, extracts
text via pdfjs-dist and rasterizes pages to images via @napi-rs/canvas,
then sends through the standard vision/text completion path
Key features:
- Single PDF (`pdf` param) or multiple PDFs (`pdfs` array, up to 10)
- Page range selection (`pages` param, e.g. "1-5", "1,3,7-9")
- Model override (`model` param) and file size limits (`maxBytesMb`)
- Auto-detects provider capability and falls back gracefully
- Same security patterns as image tool (SSRF guards, sandbox support,
local path roots, workspace-only policy)
Config (agents.defaults):
- pdfModel: primary/fallbacks (defaults to imageModel, then session model)
- pdfMaxBytesMb: max PDF file size (default: 10)
- pdfMaxPages: max pages to process (default: 20)
Model catalog:
- Extended ModelInputType to include "document" alongside "text"/"image"
- Added modelSupportsDocument() capability check
Files:
- src/agents/tools/pdf-tool.ts - main tool factory
- src/agents/tools/pdf-tool.helpers.ts - helpers (page range, config, etc.)
- src/agents/tools/pdf-native-providers.ts - direct API calls for Anthropic/Google
- src/agents/tools/pdf-tool.test.ts - 43 tests covering all paths
- Modified: model-catalog.ts, openclaw-tools.ts, config schema/types/labels/help
* fix: prepare pdf tool for merge (#31319) (thanks @tyler6204)
* feat(agents): support `thinkingDefault: "adaptive"` for Anthropic models
Anthropic's Opus 4.6 and Sonnet 4.6 support adaptive thinking where the
model dynamically decides when and how much to think. This is now
Anthropic's recommended mode and `budget_tokens` is deprecated on these
models.
Add "adaptive" as a valid thinking level:
- Config: `agents.defaults.thinkingDefault: "adaptive"`
- CLI: `/think adaptive` or `/think auto`
- Pi SDK mapping: "adaptive" → "medium" effort at the pi-agent-core
layer, which the Anthropic provider translates to
`thinking.type: "adaptive"` with `output_config.effort: "medium"`
- Provider fallbacks: OpenRouter and Google map "adaptive" to their
respective "medium" equivalents
Closes#30880
Made-with: Cursor
* style(changelog): format changelog with oxfmt
* test(types): fix strict typing in runtime/plugin-context tests
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
When sessions_spawn is called without runTimeoutSeconds, subagents
previously defaulted to 0 (no timeout). This adds a config key at
agents.defaults.subagents.runTimeoutSeconds so operators can set a
global default timeout for all subagent runs.
The agent-provided value still takes precedence when explicitly passed.
When neither the agent nor the config specifies a timeout, behavior is
unchanged (0 = no timeout), preserving backwards compatibility.
Updated for the subagent-spawn.ts refactor (logic moved from
sessions-spawn-tool.ts to spawnSubagentDirect).
Closes#19288
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The global `agents.defaults.thinkingDefault` forces a single thinking
level for all models. Users running multiple models with different
reasoning capabilities (e.g. Claude with extended thinking, GPT-4o
without, Gemini Flash with lightweight reasoning) cannot optimise the
thinking level per model.
Add an optional `thinkingDefault` field to `AgentModelEntryConfig` so
each entry under `agents.defaults.models` can declare its own default.
Resolution priority: per-model → global → catalog auto-detect.
Example config:
"models": {
"anthropic/claude-sonnet-4-20250514": { "thinkingDefault": "high" },
"openai/gpt-4o": { "thinkingDefault": "off" }
}
Co-authored-by: Cursor <cursoragent@cursor.com>
