* feat: add CI failure detail notifications in lifecycle manager
When CI fails on a PR, the lifecycle manager now fetches individual check
details (names, statuses, URLs) and sends them to the worker session.
This complements the existing static reaction message with actionable
debugging information.
Flow:
- On first transition to ci_failed: static reaction message fires (existing)
- On next poll: detailed CI failure info with check names and URLs dispatched
- Fingerprinting prevents re-sending the same failure set
- New/changed failures trigger fresh detailed notifications
- Tracking metadata cleared when PR is merged/closed or CI passes
Follows the same deduplication pattern as maybeDispatchReviewBacklog().
* fix: send CI details directly to avoid consuming reaction retry budget
The detailed CI failure dispatch now uses sessionManager.send() directly
instead of executeReaction(), so it doesn't increment the ci-failed
reaction tracker. This prevents low retries/escalateAfter settings from
causing premature escalation before the agent receives failure details.
The transition reaction still owns escalation; the detailed dispatch is
purely informational follow-up delivery.
* feat: add merge conflict notifications in lifecycle manager
Adds maybeDispatchMergeConflicts() that detects merge conflicts from
the PR enrichment cache or getMergeability() and notifies the worker
session. Conflicts are dispatched independently of session status since
they can coexist with ci_failed, changes_requested, etc.
- Uses the existing merge-conflicts reaction config
- Dispatches once per conflict occurrence (tracks lastMergeConflictDispatched)
- Clears tracking when conflicts resolve, allowing re-dispatch if they recur
- Sends directly via sessionManager.send() (same pattern as CI details)
* fix: use CICheck type instead of inline type declaration
Replace inline Array<{ name, status, url, conclusion }> with the
existing CICheck type from ./types.js for formatCIFailureMessage and
the checks variable in maybeDispatchCIFailureDetails.
* fix: resolve no-useless-assignment lint error in merge conflict check
Declare hasConflicts without initial value since both branches of the
if/else assign to it before it's read.
* feat: show agent notification state in session page blockers
The blockers section on both SessionDetail (IssuesList) and SessionCard
(alert pills) now shows whether the agent has been notified about each
blocker. Reads lifecycle manager dispatch metadata:
- lastCIFailureDispatchHash for CI failures
- lastMergeConflictDispatched for merge conflicts
- lastPendingReviewDispatchHash for review comments
Displays "agent notified" indicator next to blockers where the lifecycle
manager has already forwarded the issue to the worker session.
* fix: use lifecycle status as fallback for blockers when PR data is stale
The blockers section now uses the lifecycle manager's session status
metadata as a source of truth when PR enrichment data hasn't caught up.
PR enrichment uses a 5-min cache and can timeout or be rate-limited,
causing blockers to show stale/incorrect state.
Changes:
- IssuesList and getAlerts now check metadata["status"] (lifecycle
manager state) alongside PR enrichment data
- CI failing: shown when pr.ciStatus is "failing" OR lifecycle status
is "ci_failed"
- Changes requested: shown when pr.reviewDecision matches OR lifecycle
status is "changes_requested"
- Merge conflicts: shown when pr.mergeability.noConflicts is false OR
lifecycle dispatch metadata indicates conflicts were detected
* fix: fall back to getMergeability when cached hasConflicts is undefined
When PREnrichmentData has hasConflicts as undefined (the field is typed
as boolean | undefined), the previous check treated it as no conflicts.
Now falls through to the getMergeability() call instead.
* test: add coverage for CI/conflict notify action and recovery paths
- Test CI tracking clears when CI recovers to passing
- Test notify action for CI failure details (human notification path)
- Test notify action for merge conflicts (human notification path)
These cover the previously uncovered notify action branches and the
CI recovery cleanup path in the lifecycle manager.
* fix: resolve typecheck error in CI recovery test
writeMetadata requires SessionMetadata type which doesn't include
custom keys like lastCIFailureFingerprint. Rewrote the test to use
setupCheck and let the lifecycle manager set tracking metadata
naturally through the CI failure flow, then verify cleanup on recovery.
* fix: don't use dispatch metadata for conflict detection in UI
lastMergeConflictDispatched lingers after conflicts resolve until the
lifecycle manager's next poll clears it. Using it as a conflict signal
caused stale "merge conflict" alerts. Now only pr.mergeability.noConflicts
drives conflict detection; the metadata is only used for the "agent
notified" badge.
* fix: use Promise.allSettled for dispatch functions to avoid orphaned rejections
Promise.all rejects immediately on first failure, leaving in-flight
promises unmonitored. Promise.allSettled waits for all to complete.
|
||
|---|---|---|
| .. | ||
| __tests__ | ||
| src | ||
| CHANGELOG.md | ||
| README.md | ||
| package.json | ||
| tsconfig.build.json | ||
| tsconfig.json | ||
| vitest.config.ts | ||
README.md
@composio/ao-core
Core services, types, and configuration for the Agent Orchestrator system.
What's Here
src/types.ts— All TypeScript interfaces (Runtime, Agent, Workspace, Tracker, SCM, Notifier, Terminal, Session, events)src/services/— Core services (SessionManager, LifecycleManager, PluginRegistry)src/config.ts— Configuration loading + Zod schemassrc/utils/— Shared utilities (shell escaping, metadata parsing, etc.)
Key Files
src/types.ts — The Source of Truth
Every interface the system uses is defined here. If you're working on any part of the orchestrator, start by reading this file.
Main interfaces:
Runtime— where sessions execute (tmux, docker, k8s)Agent— AI coding tool adapter (claude-code, codex, aider)Workspace— code isolation (worktree, clone)Tracker— issue tracking (GitHub Issues, Linear)SCM— PR/CI/reviews (GitHub, GitLab)Notifier— push notifications (desktop, Slack, webhook)Terminal— human interaction UI (iTerm2, web)Session— running agent instance (state, metadata, handles)OrchestratorEvent— events emitted by lifecycle managerPluginModule— what every plugin exports
src/services/session-manager.ts — Session CRUD
Handles session lifecycle:
spawn(config)— create new session (workspace + runtime + agent)list(projectId?)— list all sessionsget(sessionId)— get session detailskill(sessionId)— terminate sessioncleanup(projectId?)— kill completed/merged sessionssend(sessionId, message)— send message to agent
Data flow in spawn():
- Load project config
- Validate issue exists via
Tracker.getIssue()(if issueId provided, fails-fast if not found) - Reserve session ID
- Determine branch name
- Create workspace via
Workspace.create() - Generate prompt via
Tracker.generatePrompt() - Build launch command via
Agent.getLaunchCommand() - Create runtime session via
Runtime.create() - Run
Agent.postLaunchSetup()(optional) - Write metadata file
- Return Session object
Note: If issue validation fails (not found, auth error), spawn fails before creating any resources (no workspace, no runtime, no session ID). This prevents spawning sessions with broken issue references.
src/services/lifecycle-manager.ts — State Machine + Reactions
Polls sessions, detects state changes, triggers reactions:
State machine:
spawning → working → pr_open → ci_failed/review_pending/approved → mergeable → merged
Reactions:
ci-failed→ send fix prompt to agentchanges-requested→ send review comments to agentapproved-and-green→ notify human (or auto-merge)agent-stuck→ notify human
Polling loop:
- For each session: check agent activity state (
Agent.getActivityState()) - If PR exists: check CI status (
SCM.getCISummary()), review state (SCM.getReviewDecision()) - Update session status based on state
- Trigger reactions if state changed
- Emit events
src/services/plugin-registry.ts — Plugin Discovery + Loading
Loads plugins and provides access to them:
register(plugin, config?)— register a plugin instanceget<T>(slot, name)— get plugin by slot + namelist(slot)— list all plugins for a slotloadBuiltins(config?)— load built-in plugins (runtime-tmux, agent-claude-code, etc.)loadFromConfig(config)— load built-ins today; external plugin descriptors are the marketplace extension point
Built-in plugins (loaded by default):
- runtime-tmux, runtime-process
- agent-claude-code, agent-codex, agent-aider, agent-opencode
- workspace-worktree, workspace-clone
- tracker-github, tracker-linear, tracker-gitlab
- scm-github, scm-gitlab
- notifier-desktop, notifier-discord, notifier-slack, notifier-composio, notifier-openclaw, notifier-webhook
- terminal-iterm2, terminal-web
src/config.ts — Configuration Loading
Loads and validates agent-orchestrator.yaml:
Main config sections:
- Runtime data paths are auto-derived from the config location under
~/.agent-orchestrator/{hash}-{projectId}/ port— web dashboard port (default 3000, set different values for multiple projects)terminalPort— terminal WebSocket port (auto-detected if not set)directTerminalPort— direct terminal WebSocket port (auto-detected if not set)defaults— default plugins (runtime, agent, workspace, notifiers)plugins— installer-managed external plugin descriptors (registry, npm, or local)projects— per-project config (repo, path, branch, symlinks, reactions, agentRules)notifiers— notification channel config (Slack webhooks, etc.)notificationRouting— which notifiers get which priority eventsreactions— auto-response config (ci-failed, changes-requested, approved-and-green, etc.)
Zod schemas validate all config at load time.
Common Tasks
Adding a Field to Session
- Edit
src/types.ts→Sessioninterface - Edit
src/services/session-manager.ts→ initialize field inspawn() - Rebuild:
pnpm --filter @composio/ao-core build
Adding an Event Type
- Edit
src/types.ts→EventTypeunion - Emit the event:
eventEmitter.emit()in relevant service - Add reaction handler (optional):
src/services/lifecycle-manager.ts
Adding a Reaction
- Edit
src/services/lifecycle-manager.ts→ add handler function - Wire it up in the polling loop
- Add config schema in
src/config.tsif new reaction type
Feedback Tools (v1)
@composio/ao-core exports two structured feedback tool contracts:
bug_reportimprovement_suggestion
Both share the same required input fields:
titlebodyevidence(array of strings)sessionsourceconfidence(0..1)
Example:
import { FEEDBACK_TOOL_NAMES, FeedbackReportStore, getFeedbackReportsDir } from "@composio/ao-core";
const reportsDir = getFeedbackReportsDir(configPath, projectPath);
const store = new FeedbackReportStore(reportsDir);
const saved = store.persist(FEEDBACK_TOOL_NAMES.BUG_REPORT, {
title: "SSO login loop",
body: "Google SSO redirects back to /login repeatedly.",
evidence: ["trace_id=abc123", "screenshot: login-loop.png"],
session: "ao-22",
source: "agent",
confidence: 0.84,
});
Storage format:
- Reports are persisted under
~/.agent-orchestrator/{hash}-{projectId}/feedback-reports - Each report is a typed key=value file (
report_<timestamp>_<id>.kv) for easy inspection - A deterministic dedupe key (
sha256, 16 hex chars) is generated from normalized tool+content
Migration notes:
- No migration needed for existing AO installs
- The
feedback-reportsdirectory is created lazily on first persisted report
Testing
# Run all core tests
pnpm --filter @composio/ao-core test
# Run in watch mode
pnpm --filter @composio/ao-core test -- --watch
# Run specific test
pnpm --filter @composio/ao-core test -- session-manager.test.ts
Tests are in src/__tests__/:
session-manager.test.ts— session CRUD, spawn, cleanuplifecycle-manager.test.ts— state machine, reactionsplugin-registry.test.ts— plugin loading, resolutiontmux.test.ts— tmux utility functions (not a plugin test)prompt-builder.test.ts— prompt generation utilities
Building
# Build core
pnpm --filter @composio/ao-core build
# Typecheck
pnpm --filter @composio/ao-core typecheck
This package is a dependency of all other packages. Build it first if working on the codebase.
Architecture Notes
Why flat metadata files?
- Debuggability:
cat ~/.agent-orchestrator/<hash>-my-app/sessions/app-3shows full state - No database dependency (survives crashes, easy to inspect)
- Backwards-compatible with bash script orchestrator
Why polling instead of webhooks?
- Simpler (no webhook setup, no ngrok for local dev)
- Works offline (CI/review state is fetched, not pushed)
- Survives orchestrator restarts (no missed events)
Why plugin slots?
- Swappability: use tmux locally, docker in CI, k8s in prod
- Testability: mock plugins for tests
- Extensibility: users can add custom plugins (e.g., company-specific notifier)