* refactor: remove channel shim directories, point all imports to extensions
Delete the 6 backward-compat shim directories (src/telegram, src/discord,
src/slack, src/signal, src/imessage, src/web) that were re-exporting from
extensions. Update all 112+ source files to import directly from
extensions/{channel}/src/ instead of through the shims.
Also:
- Move src/channels/telegram/ (allow-from, api) to extensions/telegram/src/
- Fix outbound adapters to use resolveOutboundSendDep (fixes 5 pre-existing TS errors)
- Update cross-extension imports (src/web/media.js → extensions/whatsapp/src/media.js)
- Update vitest, tsdown, knip, labeler, and script configs for new paths
- Update guard test allowlists for extension paths
After this, src/ has zero channel-specific implementation code — only the
generic plugin framework remains.
* fix: update raw-fetch guard allowlist line numbers after shim removal
* refactor: document direct extension channel imports
* test: mock transcript module in delivery helpers
Plugin subagent dispatch used a hardcoded synthetic client carrying
operator.admin, operator.approvals, and operator.pairing for all
runtime.subagent.* calls. Plugin HTTP routes with auth:"plugin" require
no gateway auth by design, so an unauthenticated external request could
drive admin-only gateway methods (sessions.delete, agent.run) through
the subagent runtime.
Propagate the real gateway client into the plugin runtime request scope
when one is available. Plugin HTTP routes now run inside a scoped
runtime client: auth:"plugin" routes receive a non-admin synthetic
operator.write client; gateway-authenticated routes retain admin-capable
scopes. The security boundary is enforced at the HTTP handler level.
Fixes GHSA-xw77-45gv-p728
* feat(context-engine): add ContextEngine interface and registry
Introduce the pluggable ContextEngine abstraction that allows external
plugins to register custom context management strategies.
- ContextEngine interface with lifecycle methods: bootstrap, ingest,
ingestBatch, afterTurn, assemble, compact, prepareSubagentSpawn,
onSubagentEnded, dispose
- Module-level singleton registry with registerContextEngine() and
resolveContextEngine() (config-driven slot selection)
- LegacyContextEngine: pass-through implementation wrapping existing
compaction behavior for 100% backward compatibility
- ensureContextEnginesInitialized() guard for safe one-time registration
- 19 tests covering contract, registry, resolution, and legacy parity
* feat(plugins): add context-engine slot and registerContextEngine API
Wire the ContextEngine abstraction into the plugin system so external
plugins can register context engines via the standard plugin API.
- Add 'context-engine' to PluginKind union type
- Add 'contextEngine' slot to PluginSlotsConfig (default: 'legacy')
- Wire registerContextEngine() through OpenClawPluginApi
- Export ContextEngine types from plugin-sdk for external consumers
- Restore proper slot-based resolution in registry
* feat(context-engine): wire ContextEngine into agent run lifecycle
Integrate the ContextEngine abstraction into the core agent run path:
- Resolve context engine once per run (reused across retries)
- Bootstrap: hydrate canonical store from session file on first run
- Assemble: route context assembly through pluggable engine
- Auto-compaction guard: disable built-in auto-compaction when
the engine declares ownsCompaction (prevents double-compaction)
- AfterTurn: post-turn lifecycle hook for ingest + background
compaction decisions
- Overflow compaction: route through contextEngine.compact()
- Dispose: clean up engine resources in finally block
- Notify context engine on subagent lifecycle events
Legacy engine: all lifecycle methods are pass-through/no-op, preserving
100% backward compatibility for users without a context engine plugin.
* feat(plugins): add scoped subagent methods and gateway request scope
Expose runtime.subagent.{run, waitForRun, getSession, deleteSession}
so external plugins can spawn sub-agent sessions without raw gateway
dispatch access.
Uses AsyncLocalStorage request-scope bridge to dispatch internally via
handleGatewayRequest with a synthetic operator client. Methods are only
available during gateway request handling.
- Symbol.for-backed global singleton for cross-module-reload safety
- Fallback gateway context for non-WS dispatch paths (Telegram/WhatsApp)
- Set gateway request scope for all handlers, not just plugin handlers
- 3 staleness tests for fallback context hardening
* feat(context-engine): route /compact and sessions.get through context engine
Wire the /compact command and sessions.get handler through the pluggable
ContextEngine interface.
- Thread tokenBudget and force parameters to context engine compact
- Route /compact through contextEngine.compact() when registered
- Wire sessions.get as runtime alias for plugin subagent dispatch
- Add .pebbles/ to .gitignore
* style: format with oxfmt 0.33.0
Fix duplicate import (ControlUiRootState in server.impl.ts) and
import ordering across all changed files.
* fix: update extension test mocks for context-engine types
Add missing subagent property to bluebubbles PluginRuntime mock.
Add missing registerContextEngine to lobster OpenClawPluginApi mock.
* fix(subagents): keep deferred delete cleanup retryable
* style: format run attempt for CI
* fix(rebase): remove duplicate embedded-run imports
* test: add missing gateway context mock export
* fix: pass resolved auth profile into afterTurn compaction
Ensure the embedded runner forwards resolved auth profile context into
legacy context-engine compaction params on the normal afterTurn path,
matching overflow compaction behavior. This allows downstream LCM
summarization to use the intended provider auth/profile consistently.
Also fix strict TS typing in external-link token dedupe and align an
attempt unit test reasoningLevel value with the current ReasoningLevel
enum.
Regeneration-Prompt: |
We were debugging context-engine compaction where downstream summary
calls were missing the right auth/profile context in normal afterTurn
flow, while overflow compaction already propagated it. Preserve current
behavior and keep changes additive: thread the resolved authProfileId
through run -> attempt -> legacy compaction param builder without
broad refactors.
Add tests that prove the auth profile is included in afterTurn legacy
params and that overflow compaction still passes it through run
attempts. Keep existing APIs stable, and only adjust small type issues
needed for strict compilation.
* fix: remove duplicate imports from rebase
* feat: add context-engine system prompt additions
* fix(rebase): dedupe attempt import declarations
* test: fix fetch mock typing in ollama autodiscovery
* fix(test): add registerContextEngine to diffs extension mock APIs
* test(windows): use path.delimiter in ios-team-id fixture PATH
* test(cron): add model formatting and precedence edge case tests
Covers:
- Provider/model string splitting (whitespace, nested paths, empty segments)
- Provider normalization (casing, aliases like bedrock→amazon-bedrock)
- Anthropic model alias normalization (opus-4.5→claude-opus-4-5)
- Precedence: job payload > session override > config default
- Sequential runs with different providers (CI flake regression pattern)
- forceNew session preserving stored model overrides
- Whitespace/empty model string edge cases
- Config model as string vs object format
* test(cron): fix model formatting test config types
* test(phone-control): add registerContextEngine to mock API
* fix: re-export ChannelKind from config-reload-plan
* fix: add subagent mock to plugin-runtime-mock test util
* docs: add changelog fragment for context engine PR #22201
Add requestHeartbeatNow to PluginRuntime.system so extensions can
trigger an immediate heartbeat wake without importing internal modules.
This enables extensions to inject a system event and wake the agent
in one step — useful for inbound message handlers that use the
heartbeat model (e.g. agent-to-agent DMs via Nostr).
Changes:
- src/plugins/runtime/types.ts: add RequestHeartbeatNow type alias
and requestHeartbeatNow to PluginRuntime.system
- src/plugins/runtime/index.ts: import and wire requestHeartbeatNow
into createPluginRuntime()
* feat(hooks): add trigger and channelId to plugin hook agent context
Adds `trigger` and `channelId` fields to `PluginHookAgentContext` so
plugins can determine what initiated the agent run and which channel
it originated from, without session-key parsing or Redis bridging.
trigger values: "user", "heartbeat", "cron", "memory"
channelId values: "telegram", "discord", "whatsapp", etc.
Both fields are threaded through run.ts and attempt.ts hookCtx so all
hook phases receive them (before_model_resolve, before_prompt_build,
before_agent_start, llm_input, llm_output, agent_end).
channelId falls back from messageChannel to messageProvider when the
former is not set. followup-runner passes originatingChannel so queued
followup runs also carry channel context.
* docs(changelog): note hook context parity fix for #28623
---------
Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
* fix(plugins): expose ephemeral sessionId in tool contexts for per-conversation isolation
The plugin tool context (`OpenClawPluginToolContext`) and tool hook
context (`PluginHookToolContext`) only provided `sessionKey`, which
is a durable channel identifier that survives /new and /reset.
Plugins like mem0 that need per-conversation isolation (e.g. mapping
Mem0 `run_id`) had no way to distinguish between conversations,
causing session-scoped memories to persist unbounded across resets.
Add `sessionId` (ephemeral UUID regenerated on /new and /reset) to:
- `OpenClawPluginToolContext` (factory context for plugin tools)
- `PluginHookToolContext` (before_tool_call / after_tool_call hooks)
- Internal `HookContext` for tool call wrappers
Thread the value from the run attempt through createOpenClawCodingTools
→ createOpenClawTools → resolvePluginTools and through the tool hook
wrapper.
Closes#31253
Made-with: Cursor
* fix(agents): propagate embedded sessionId through tool hook context
* test(hooks): cover sessionId in embedded tool hook contexts
* docs(changelog): add sessionId hook context follow-up note
* test(hooks): avoid toolCallId collision in after_tool_call e2e
---------
Co-authored-by: SidQin-cyber <sidqin0410@gmail.com>
* fix(hooks): deduplicate after_tool_call hook in embedded runs
(cherry picked from commit c129a1a74b)
* fix(hooks): propagate sessionKey in after_tool_call context
The after_tool_call hook in handleToolExecutionEnd was passing
`sessionKey: undefined` in the ToolContext, even though the value is
available on ctx.params. This broke plugins that need session context
in after_tool_call handlers (e.g., for per-session audit trails or
security logging).
- Add `sessionKey` to the `ToolHandlerParams` Pick type
- Pass `ctx.params.sessionKey` through to the hook context
- Add test assertion to prevent regression
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
(cherry picked from commit b7117384fc)
* fix(hooks): thread agentId through to after_tool_call hook context
Follow-up to #30511 — the after_tool_call hook context was passing
`agentId: undefined` because SubscribeEmbeddedPiSessionParams did not
carry the agent identity. This threads sessionAgentId (resolved in
attempt.ts) through the session params into the tool handler context,
giving plugins accurate agent-scoped context for both before_tool_call
and after_tool_call hooks.
Changes:
- Add `agentId?: string` to SubscribeEmbeddedPiSessionParams
- Add "agentId" to ToolHandlerParams Pick type
- Pass `agentId: sessionAgentId` at the subscribeEmbeddedPiSession()
call site in attempt.ts
- Wire ctx.params.agentId into the after_tool_call hook context
- Update tests to assert agentId propagation
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
(cherry picked from commit aad01edd3e)
* changelog: credit after_tool_call hook contributors
* Update CHANGELOG.md
* agents: preserve adjusted params until tool end
* agents: emit after_tool_call with adjusted args
* tests: cover adjusted after_tool_call params
* tests: align adapter after_tool_call expectation
---------
Co-authored-by: jbeno <jim@jimbeno.net>
Co-authored-by: scoootscooob <zhentongfan@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Expose audio transcription through the PluginRuntime so external
plugins (e.g. marmot) can use openclaw's media-understanding provider
framework without importing unexported internal modules.
The new transcribeAudioFile() wraps runCapability({capability: "audio"})
and reads provider/model/apiKey from tools.media.audio in the config,
matching the pattern used by the Discord VC implementation.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When `openclaw plugins install diffs` downloads the unrelated npm
package `diffs@0.1.1` (which lacks `openclaw.extensions`), the install
fails without trying the bundled `@openclaw/diffs` plugin.
Two fixes:
1. Broaden the bundled-fallback trigger to also fire on
"missing openclaw.extensions" errors (not just npm 404s)
2. Match bundled plugins by pluginId in addition to npmSpec so
unscoped names like "diffs" resolve to `@openclaw/diffs`
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(plugins): fallback install entrypoints for legacy manifests
* Voice Call: enforce exact webhook path match
* Tests: isolate webhook path suite and reset cron auth state
* chore: keep #31930 scoped to voice webhook path fix
* fix: add changelog for exact voice webhook path match (#31930) (thanks @afurm)
* fix: handle HTTP 529 (Anthropic overloaded) in failover error classification
Classify Anthropic's 529 status code as "rate_limit" so model fallback
triggers reliably without depending on fragile message-based detection.
Closes#28502
* fix: add changelog for HTTP 529 failover classification (#31854) (thanks @bugkill3r)
* fix(slack): guard against undefined text in includes calls during mention handling
* fix: add changelog for mentions/slack null-safe guards (#31865) (thanks @stone-jin)
* fix(memory-lancedb): pass dimensions to embedding API call
- Add dimensions parameter to Embeddings constructor
- Pass dimensions to OpenAI embeddings.create() API call
- Fixes dimension mismatch when using custom embedding models like DashScope text-embedding-v4
* fix: add regression for memory-lancedb dimensions pass-through (#32036) (thanks @scotthuang)
* fix(telegram): guard malformed native menu specs
* fix: harden plugin command registration + telegram menu guard (#31997) (thanks @liuxiaopai-ai)
* fix(gateway): restart heartbeat on model config changes
* fix: add changelog credit for heartbeat model reload (#32046) (thanks @stakeswky)
* test(process): replace no-output timer subprocess with spawn mock
* test(perf): trim repeated setup in cron memory and config suites
* test(perf): reduce per-case setup in script and git-hook tests
* fix(slack): scope debounce key by message timestamp to prevent cross-thread collisions
Top-level channel messages from the same sender shared a bare channel
debounce key, causing concurrent messages in different threads to merge
into a single reply on the wrong thread. Now the debounce key includes
the message timestamp for top-level messages, matching how the downstream
session layer already scopes by canonicalThreadId.
Extracted buildSlackDebounceKey() for testability.
Closes#31935
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: harden slack debounce key routing and ordering (#31951) (thanks @scoootscooob)
* fix(openrouter): skip reasoning.effort injection for x-ai/grok models
x-ai/grok models on OpenRouter do not support the reasoning.effort
parameter and reject payloads containing it with "Invalid arguments
passed to the model." Skip reasoning injection for these models, the
same way we already skip it for the dynamic "auto" routing model.
Closes#32039
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: add changelog credit for openrouter x-ai reasoning guard (#32054) (thanks @scoootscooob)
* fix(agents): scope volcengine-plan/byteplus-plan auth lookup to profile resolution
The configure flow stores auth credentials under `provider: "volcengine"`,
but the coding model uses `volcengine-plan` as its provider. Add a scoped
`normalizeProviderIdForAuth` function used only by `listProfilesForProvider`
so coding-plan variants resolve to their base provider for auth credential
lookup without affecting global provider routing.
Closes#31731
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(tools): honor fsPolicy.workspaceOnly in image/pdf tool localRoots
PR #28822 fixed the Write/Edit tools to respect `tools.fs.workspaceOnly`,
but the image and PDF tools still unconditionally include default local
roots (`~/.openclaw/media`, `~/.openclaw/agents`, etc.) when computing
the `localRoots` allowlist for non-sandbox mode.
When `fsPolicy.workspaceOnly` is true, restrict `localRoots` to only the
workspace directory so that files outside the workspace are rejected by
`assertLocalMediaAllowed()`.
Relates to #31716
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: add changelog credit for fsPolicy image/pdf propagation (#31882) (thanks @justinhuangcode)
* fix: skip Telegram command sync when menu is unchanged (#32017)
Hash the command list and cache it to disk per account. On restart,
compare the current hash against the cached one and skip the
deleteMyCommands + setMyCommands round-trip when nothing changed.
This prevents 429 rate-limit errors when the gateway restarts
several times in quick succession.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(telegram): scope command-sync hash cache by bot identity (#32059)
* fix: normalize coding-plan providers in auth order validation
* feat(security): Harden Docker browser container chromium flags (#23889) (#31504)
* Gateway: honor OPENCLAW_GATEWAY_URL override for remote/local calls
* Agents: fix sandbox sessionKey usage for PI embedded subagent calls
* Sandbox: tighten browser container Chromium runtime flags
* fix: add sandbox browser defaults for container hardening
* docs: expand sandbox browser default flags list
* fix: make sandbox browser flags optional and preserve gateway env auth overrides
* docs: scope PR 31504 changelog entry
* style: format gateway call override handling
* fix: dedupe sandbox browser chrome args
* fix: preserve remote tls fingerprint for env gateway override
* fix: enforce auth for env gateway URL override
* chore: document gateway override auth security expectations
* fix(delivery): strip HTML tags for plain-text messaging surfaces
Models occasionally produce HTML tags in their output. While these render
fine on web surfaces, they appear as literal text on WhatsApp, Signal,
SMS, IRC, and Telegram.
Add sanitizeForPlainText() utility that converts common inline HTML to
lightweight-markup equivalents and strips remaining tags. Applied in the
outbound delivery pipeline for non-HTML surfaces only.
Closes#31884
See also: #18558
* fix(outbound): harden plain-text HTML sanitization paths (#32034)
* fix(security): harden file installs and race-path tests
* matrix: bootstrap crypto runtime when npm scripts are skipped
* fix(matrix): keep plugin register sync while bootstrapping crypto runtime (#31989)
* perf(runtime): reduce cron persistence and logger overhead
* test(perf): use prebuilt plugin install archive fixtures
* test(perf): increase guardrail scan read concurrency
* fix(queue): restart drain when message enqueued after idle window
After a drain loop empties the queue it deletes the key from
FOLLOWUP_QUEUES. If a new message arrives at that moment
enqueueFollowupRun creates a fresh queue object with draining:false
but never starts a drain, leaving the message stranded until the
next run completes and calls finalizeWithFollowup.
Fix: persist the most recent runFollowup callback per queue key in
FOLLOWUP_RUN_CALLBACKS (drain.ts). enqueueFollowupRun now calls
kickFollowupDrainIfIdle after a successful push; if a cached
callback exists and no drain is running it calls scheduleFollowupDrain
to restart immediately. clearSessionQueues cleans up the callback
cache alongside the queue state.
* fix: avoid stale followup drain callbacks (#31902) (thanks @Lanfei)
* fix(synology-chat): read cfg from outbound context so incomingUrl resolves
* fix: require openclaw.extensions for plugin installs (#32055) (thanks @liuxiaopai-ai)
---------
Co-authored-by: Andrii Furmanets <furmanets.andriy@gmail.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Co-authored-by: Saurabh <skmishra1991@gmail.com>
Co-authored-by: stone-jin <1520006273@qq.com>
Co-authored-by: scotthuang <scotthuang@tencent.com>
Co-authored-by: User <user@example.com>
Co-authored-by: scoootscooob <zhentongfan@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: justinhuangcode <justinhuangcode@users.noreply.github.com>
Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
Co-authored-by: AytuncYildizli <cryptosquanch@gmail.com>
Co-authored-by: bmendonca3 <bmendonca3@users.noreply.github.com>
Co-authored-by: Jealous <CooLanfei@163.com>
Co-authored-by: white-rm <zhang.xujin@xydigit.com>
The handleAutoCompactionStart handler was calling runBeforeCompaction with
only messageCount and an empty hook context. Plugins receiving this hook
could not identify the session or snapshot the transcript during
auto-compaction.
The other call site in compact.ts already passes the full payload
(messages, sessionFile, sessionKey). This aligns the subscribe handler
to do the same using ctx.params.session and ctx.params.sessionKey.
(cherry picked from commit 318a19d1a1)
* fix(plugins): use manifest id as config key instead of npm package name
Plugin manifests (openclaw.plugin.json) define a canonical 'id' field that
is used as the authoritative plugin identifier by the manifest registry.
However, the install command was deriving the config entry key from the npm
package name (e.g. 'cognee-openclaw') rather than the manifest id (e.g.
'memory-cognee'), causing a latent mismatch.
On the next gateway reload the plugin could not be found under the config key
derived from the npm package name, causing 'plugin not found' errors and
potentially shutting the gateway down.
Fix: after extracting the package directory, read openclaw.plugin.json and
prefer its 'id' field over the npm package name when registering the config
entry. Falls back to the npm-derived id if the manifest file is absent or
has no valid id. A diagnostic info message is emitted when the two values
differ so the mismatch is visible in the install log.
The update path (src/plugins/update.ts) already correctly reads the manifest
id and is unaffected.
Fixes#24429
* fix: format plugin install manifest-id path (#24796)
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* Config UI: add tag filters and complete schema help/labels
* Config UI: finalize tags/help polish and unblock test suite
* Protocol: regenerate Swift gateway models
Synchronous hook that lets plugins inspect and optionally block messages
before they are written to the session JSONL file. Primary use case is
private mode... when enabled, the plugin returns { block: true } and the
message never gets persisted.
The hook runs on the hot path (synchronous, like tool_result_persist).
Handlers execute sequentially in priority order. If any handler returns
{ block: true }, the write is skipped immediately. Handlers can also
return a modified message to write instead of the original.
Changes:
- src/plugins/types.ts: add hook name, event/result types, handler map entry
- src/plugins/hooks.ts: add runBeforeMessageWrite() following tool_result_persist pattern
- src/agents/session-tool-result-guard.ts: invoke hook before every originalAppend() call
- src/agents/session-tool-result-guard-wrapper.ts: wire hook runner to the guard
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds a complete Linq iMessage channel adapter that replaces the existing
iMessage channel's Mac Mini + dedicated Apple ID + SSH wrapper + Full Disk
Access setup with a single API key and phone number.
Core implementation (src/linq/):
- types.ts: Linq webhook event and message types
- accounts.ts: Multi-account resolution from config (env/file/inline token)
- send.ts: REST outbound via Linq Blue V3 API (messages, typing, reactions)
- probe.ts: Health check via GET /v3/phonenumbers
- monitor.ts: Webhook HTTP server with HMAC-SHA256 signature verification,
replay protection, inbound debouncing, and full dispatch pipeline integration
Extension plugin (extensions/linq/):
- ChannelPlugin implementation with config, security, setup, outbound,
gateway, and status adapters
- Supports direct and group chats, reactions, and media
Wiring:
- Channel registry, dock, config schema, plugin-sdk exports, and plugin
runtime all updated to include the new linq channel
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Pre-existing test mocks lacked pendingMessagingMediaUrls and
messagingToolSentMediaUrls fields added by the media dedup feature,
causing runtime errors in handleToolExecutionEnd.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Make before_agent_start override merging preserve the first defined
model/provider override so higher-priority hooks cannot be overwritten by
lower-priority handlers, and align the corresponding test title and
expectation with the intended precedence behavior.
Co-authored-by: Cursor <cursoragent@cursor.com>
Layer 1: Hook merger tests verify modelOverride/providerOverride are
correctly propagated through the before_agent_start merger with
priority ordering, backward compatibility, and field isolation.
Layer 2: Pipeline wiring tests verify the earlyHookResult passthrough
contract between run.ts and attempt.ts, graceful error degradation,
and that overrides correctly modify provider/model variables.
19 tests total across 2 test files.
Enable plugins to override the model and provider for agent runs by
returning modelOverride/providerOverride from the before_agent_start
hook. The hook is now invoked early in run.ts (before resolveModel)
so overrides take effect. The result is passed to attempt.ts via
earlyHookResult to prevent double-firing.
This enables security-critical use cases like routing PII-containing
prompts to local models instead of cloud providers.
* Agents: add subagent orchestration controls
* Agents: add subagent orchestration controls (WIP uncommitted changes)
* feat(subagents): add depth-based spawn gating for sub-sub-agents
* feat(subagents): tool policy, registry, and announce chain for nested agents
* feat(subagents): system prompt, docs, changelog for nested sub-agents
* fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback
Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex),
the fallback candidate logic in resolveFallbackCandidates silently appended the
global primary model (opus) as a backstop. On reinjection/steer with a transient
error, the session could fall back to opus which has a smaller context window
and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? []
instead of undefined, preventing the implicit primary backstop.
Bug 2: Active subagents showed 'model n/a' in /subagents list because
resolveModelDisplay only read entry.model/modelProvider (populated after run
completes). Fix: fall back to modelOverride/providerOverride fields which are
populated at spawn time via sessions.patch.
Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could
theoretically escape runEmbeddedPiAgent and be treated as failover candidates
in runWithModelFallback, causing a switch to a model with a smaller context
window. Fix: in runWithModelFallback, detect context overflow errors via
isLikelyContextOverflowError and rethrow them immediately instead of trying the
next model candidate.
* fix(subagents): track spawn depth in session store and fix announce routing for nested agents
* Fix compaction status tracking and dedupe overflow compaction triggers
* fix(subagents): enforce depth block via session store and implement cascade kill
* fix: inject group chat context into system prompt
* fix(subagents): always write model to session store at spawn time
* Preserve spawnDepth when agent handler rewrites session entry
* fix(subagents): suppress announce on steer-restart
* fix(subagents): fallback spawned session model to runtime default
* fix(subagents): enforce spawn depth when caller key resolves by sessionId
* feat(subagents): implement active-first ordering for numeric targets and enhance task display
- Added a test to verify that subagents with numeric targets follow an active-first list ordering.
- Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity.
- Enhanced task display in command responses to prevent truncation of long task descriptions.
- Introduced new utility functions for compacting task text and managing subagent run states.
* fix(subagents): show model for active runs via run record fallback
When the spawned model matches the agent's default model, the session
store's override fields are intentionally cleared (isDefault: true).
The model/modelProvider fields are only populated after the run
completes. This left active subagents showing 'model n/a'.
Fix: store the resolved model on SubagentRunRecord at registration
time, and use it as a fallback in both display paths (subagents tool
and /subagents command) when the session store entry has no model info.
Changes:
- SubagentRunRecord: add optional model field
- registerSubagentRun: accept and persist model param
- sessions-spawn-tool: pass resolvedModel to registerSubagentRun
- subagents-tool: pass run record model as fallback to resolveModelDisplay
- commands-subagents: pass run record model as fallback to resolveModelDisplay
* feat(chat): implement session key resolution and reset on sidebar navigation
- Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar.
- Updated the `renderTab` function to handle session key changes when navigating to the chat tab.
- Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation.
* fix: subagent timeout=0 passthrough and fallback prompt duplication
Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default
- sessions-spawn-tool: default to undefined (not 0) when neither timeout param
is provided; use != null check so explicit 0 passes through to gateway
- agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles
0 → MAX_SAFE_TIMEOUT_MS)
Bug 2: model fallback no longer re-injects the original prompt as a duplicate
- agent.ts: track fallback attempt index; on retries use a short continuation
message instead of the full original prompt since the session file already
contains it from the first attempt
- Also skip re-sending images on fallback retries (already in session)
* feat(subagents): truncate long task descriptions in subagents command output
- Introduced a new utility function to format task previews, limiting their length to improve readability.
- Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately.
- Adjusted related tests to verify that long task descriptions are now truncated in the output.
* refactor(subagents): update subagent registry path resolution and improve command output formatting
- Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically.
- Enhanced the formatting of command output for active and recent subagents, adding separators for better readability.
- Updated related tests to reflect changes in command output structure.
* fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted
The previous fix (75a791106) correctly handled the case where
runTimeoutSeconds was explicitly set to 0 ("no timeout"). However,
when models omit the parameter entirely (which is common since the
schema marks it as optional), runTimeoutSeconds resolved to undefined.
undefined flowed through the chain as:
sessions_spawn → timeout: undefined (since undefined != null is false)
→ gateway agent handler → agentCommand opts.timeout: undefined
→ resolveAgentTimeoutMs({ overrideSeconds: undefined })
→ DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes)
This caused subagents to be killed at exactly 10 minutes even though
the user's intent (via TOOLS.md) was for subagents to run without a
timeout.
Fix: default runTimeoutSeconds to 0 (no timeout) when neither
runTimeoutSeconds nor timeoutSeconds is provided by the caller.
Subagent spawns are long-running by design and should not inherit the
600s agent-command default timeout.
* fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default)
* fix: thread timeout override through getReplyFromConfig dispatch path
getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override,
always falling back to the config default (600s). Add timeoutOverrideSeconds
to GetReplyOptions and pass it through as overrideSeconds so callers of the
dispatch chain can specify a custom timeout (0 = no timeout).
This complements the existing timeout threading in agentCommand and the
cron isolated-agent runner, which already pass overrideSeconds correctly.
* feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling
- Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution.
- Updated the `resolveFallbackCandidates` function to utilize the new normalization logic.
- Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms.
- Introduced a new test case to ensure that the normalization process works as expected for various input formats.
* feat(tests): add unit tests for steer failure behavior in openclaw-tools
- Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails.
- Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected.
- Enhanced the subagent registry with a new function to clear steer restart suppression.
- Updated related components to support the new test scenarios.
* fix(subagents): replace stop command with kill in slash commands and documentation
- Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs.
- Modified related documentation to reflect the change in command usage.
- Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling.
- Enhanced tests to ensure correct behavior of the updated commands and their interactions.
* feat(tests): add unit tests for readLatestAssistantReply function
- Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios.
- Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text.
- Mocked the gateway call to simulate different message histories for comprehensive testing.
* feat(tests): enhance subagent kill-all cascade tests and announce formatting
- Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents.
- Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content.
- Improved the handling of long findings and stats in the announce formatting logic to ensure concise output.
- Refactored related functions to enhance clarity and maintainability in the subagent registry and tools.
* refactor(subagent): update announce formatting and remove unused constants
- Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests.
- Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic.
- Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs.
- Cleaned up unused imports in the commands-subagents file to enhance code clarity.
* feat(tests): enhance billing error handling in user-facing text
- Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context.
- Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages.
- Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output.
- Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements.
* feat(subagent): enhance workflow guidance and auto-announcement clarity
- Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates.
- Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow.
- Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts.
* fix(cron): avoid announcing interim subagent spawn acks
* chore: clean post-rebase imports
* fix(cron): fall back to child replies when parent stays interim
* fix(subagents): make active-run guidance advisory
* fix(subagents): update announce flow to handle active descendants and enhance test coverage
- Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting.
- Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents.
- Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process.
* fix(subagents): enhance announce flow and formatting for user updates
- Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context.
- Refactored the announcement logic to improve clarity and ensure internal context remains private.
- Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates.
- Introduced a new function to build reply instructions based on session context, improving the overall announcement process.
* fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204)
* fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204)
* fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204)
* fix: apply formatting after rebase (#14447) (thanks @tyler6204)
When the same plugin directory is discovered through different path
representations (e.g. symlinks), the manifest registry incorrectly
warns about a duplicate plugin id. This is a false positive that
appears for bundled extensions like feishu (#16208).
Compare fs.realpathSync() of both candidates' rootDir before emitting
the duplicate warning. If they resolve to the same physical directory,
silently skip the duplicate instead of warning.
Also change seenIds from Set<string> to Map<string, PluginCandidate>
to track the first-seen candidate for comparison.
Closes#16208
* feat: add before_compaction and before_reset plugin hooks with session context
- Pass session messages to before_compaction hook
- Add before_reset plugin hook for /new and /reset commands
- Add sessionId to plugin hook agent context
* feat: extraBootstrapFiles config with glob pattern support
Add extraBootstrapFiles to agent defaults config, allowing glob patterns
(e.g. "projects/*/TOOLS.md") to auto-load project-level bootstrap files
into agent context every turn. Missing files silently skipped.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix(status): show custom memory plugins as enabled, not unavailable
The status command probes memory availability using the built-in
memory-core manager. Custom memory plugins (e.g. via plugin slot)
can't be probed this way, so they incorrectly showed "unavailable".
Now they show "enabled (plugin X)" without the misleading label.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: use async fs.glob and capture pre-compaction messages
- Replace globSync (node:fs) with fs.glob (node:fs/promises) to match
codebase conventions for async file operations
- Capture session.messages BEFORE replaceMessages(limited) so
before_compaction hook receives the full conversation history,
not the already-truncated list
* fix: resolve lint errors from CI (oxlint strict mode)
- Add void to fire-and-forget IIFE (no-floating-promises)
- Use String() for unknown catch params in template literals
- Add curly braces to single-statement if (curly rule)
* fix: resolve remaining CI lint errors in workspace.ts
- Remove `| string` from WorkspaceBootstrapFileName union (made all
typeof members redundant per no-redundant-type-constituents)
- Use type assertion for extra bootstrap file names
- Drop redundant await on fs.glob() AsyncIterable (await-thenable)
* fix: address Greptile review — path traversal guard + fs/promises import
- workspace.ts: use path.resolve() + traversal check in loadExtraBootstrapFiles()
- commands-core.ts: import fs from node:fs/promises, drop fs.promises prefix
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: resolve symlinks before workspace boundary check
Greptile correctly identified that symlinks inside the workspace could
point to files outside it, bypassing the path prefix check. Now uses
fs.realpath() to resolve symlinks before verifying the real path stays
within the workspace boundary.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address Greptile review — hook reliability and type safety
1. before_compaction: add compactingCount field so plugins know both
the full pre-compaction message count and the truncated count being
fed to the compaction LLM. Clarify semantics in comment.
2. loadExtraBootstrapFiles: use path.basename() for the name field
so "projects/quaid/TOOLS.md" maps to the known "TOOLS.md" type
instead of an invalid WorkspaceBootstrapFileName cast.
3. before_reset: fire the hook even when no session file exists.
Previously, short sessions without a persisted file would silently
skip the hook. Now fires with empty messages array so plugins
always know a reset occurred.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: validate bootstrap filenames and add compaction hook timeout
- Only load extra bootstrap files whose basename matches a recognized
workspace filename (AGENTS.md, TOOLS.md, etc.), preventing arbitrary
files from being injected into agent context.
- Wrap before_compaction hook in a 30-second Promise.race timeout so
misbehaving plugins cannot stall the compaction pipeline.
- Clarify hook comments: before_compaction is intentionally awaited
(plugins need messages before they're discarded) but bounded.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: make before_compaction non-blocking, add sessionFile to after_compaction
- before_compaction is now true fire-and-forget — no await, no timeout.
Plugins that need full conversation data should persist it themselves
and return quickly, or use after_compaction for async processing.
- after_compaction now includes sessionFile path so plugins can read
the full JSONL transcript asynchronously. All pre-compaction messages
are preserved on disk, eliminating the need to block compaction.
- Removes Promise.race timeout pattern that didn't actually cancel
slow hooks (just raced past them while they continued running).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add sessionFile to before_compaction for parallel processing
The session JSONL already has all messages on disk before compaction
starts. By providing sessionFile in before_compaction, plugins can
read and extract data in parallel with the compaction LLM call rather
than waiting for after_compaction. This is the optimal path for memory
plugins that need the full conversation history.
sessionFile is also kept on after_compaction for plugins that only
need to act after compaction completes (analytics, cleanup, etc.).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: move bootstrap extras into bundled hook
---------
Co-authored-by: Solomon Steadman <solstead@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Clawdbot <clawdbot@alfie.local>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* TypeScript: add extensions to tsconfig and fix type errors
- Add extensions/**/* to tsconfig.json includes
- Export ProviderAuthResult, AnyAgentTool from plugin-sdk
- Fix optional chaining for messageActions across channels
- Add missing type imports (MSTeamsConfig, GroupPolicy, etc.)
- Add type annotations for provider auth handlers
- Fix undici/fetch type compatibility in zalo proxy
- Correct ChannelAccountSnapshot property usage
- Add type casts for tool registrations
- Extract usage view styles and types to separate files
* TypeScript: fix optional debug calls and handleAction guards
The previous migration to tsdown was reverted because it caused a ~20x slowdown when running OpenClaw from the repo. @hyf0 investigated and found that simply renaming the `dist` folder also caused the same slowdown. It turns out the Plugin script loader has a bunch of voodoo vibe logic to determine if it should load files from source and compile them, or if it should load them from dist. When building with tsdown, the filesystem layout is different (bundled), and so some files weren't in the right location, and the Plugin script loader decided to compile source files from scratch using Jiti.
The new implementation uses tsdown to embed `NODE_ENV: 'production'`, which we now use to determine if we are running OpenClaw from a "production environmen" (ie. from dist). This removes the slop in favor of a deterministic toggle, and doesn't rely on directory names or similar.
There is some code reaching into `dist` to load specific modules, primarily in the voice-call extension, which I simplified into loading an "officially" exported `extensionAPI.js` file. With tsdown, entry points need to be explicitly configured, so we should be able to avoid sloppy code reaching into internals from now on. This might break some existing users, but if it does, it's because they were using "private" APIs.
* feat: add chunking mode for outbound messages
- Introduced `chunkMode` option in various account configurations to allow splitting messages by "length" or "newline".
- Updated message processing to handle chunking based on the selected mode.
- Added tests for new chunking functionality, ensuring correct behavior for both modes.
* feat: enhance chunking mode documentation and configuration
- Added `chunkMode` option to the BlueBubbles account configuration, allowing users to choose between "length" and "newline" for message chunking.
- Updated documentation to clarify the behavior of the `chunkMode` setting.
- Adjusted account merging logic to incorporate the new `chunkMode` configuration.
* refactor: simplify chunk mode handling for BlueBubbles
- Removed `chunkMode` configuration from various account schemas and types, centralizing chunk mode logic to BlueBubbles only.
- Updated `processMessage` to default to "newline" for BlueBubbles chunking.
- Adjusted tests to reflect changes in chunk mode handling for BlueBubbles, ensuring proper functionality.
* fix: update default chunk mode to 'length' for BlueBubbles
- Changed the default value of `chunkMode` from 'newline' to 'length' in the BlueBubbles configuration and related processing functions.
- Updated documentation to reflect the new default behavior for chunking messages.
- Adjusted tests to ensure the correct default value is returned for BlueBubbles chunk mode.
- Add registry lock during command execution to prevent race conditions
- Add input sanitization for command arguments (defense in depth)
- Validate handler is a function during registration
- Remove redundant case-insensitive regex flag
- Add success logging for command execution
- Simplify handler return type (always returns result now)
- Remove dead code branch in commands-plugin.ts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add clearPluginCommands() call in loadClawdbotPlugins() to ensure
previously registered commands are cleaned up before reloading plugins.
This prevents command conflicts during hot-reload scenarios.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Blockers fixed:
- Fix documentation: requireAuth defaults to true (not false)
- Add command name validation (must start with letter, alphanumeric only)
- Add reserved commands list to prevent shadowing built-in commands
- Emit diagnostic errors for invalid/duplicate command registration
Other improvements:
- Return user-friendly message for unauthorized commands (instead of silence)
- Sanitize error messages to avoid leaking internal details
- Document acceptsArgs behavior when arguments are provided
- Add notes about reserved commands and validation rules to docs
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This adds a new `api.registerCommand()` method to the plugin API, allowing
plugins to register slash commands that execute without invoking the AI agent.
Features:
- Plugin commands are processed before built-in commands and the agent
- Commands can optionally require authorization
- Commands can accept arguments
- Async handlers are supported
Use case: plugins can implement toggle commands (like /tts_on, /tts_off)
that respond immediately without consuming LLM API calls.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Harold Hunt
Peter Steinberger
Gustavo Madeira Santana
Nimrod Gutman
Vincent Koc
xiaoyi
Tak Hoffman
ufhy
scoootscooob
Robin Waslander
Xinhua Gu
Josh Lehman
maweibin
Igal Tabachnik
Cui Chen
Shadow
Muhammed Mukhthar CM
tempeste
Austin Eral
David Rudduck
SciFantastic
markfietje
benthecarman
scoootscooob
Mark L
Shakker
Onur Solmaz
Marc Gratch
zerone0x
chilu18
Vignesh Natarajan
Onur
Mariano
Sebastian
cpojer
Parker Todd Brooks
George McCain
Yaroslav Boiko
Nate Fikru
David Harmeyer
Tyler Yust
Robby
Peter Steinberger
shadril238
solstead
Omair Afzal
JustasM
Patrick Barletta
Kyle Tse
max
quotentiroler
Mariano Belinky
Abdel Sy Fane
Josh Palmer
plum-dawg
Glucksberg
André Abadesso
Shadow
Doug von Kohorn
Sebastian Slight
Radek Paclt