* feat: implement notifier and terminal plugins (desktop, slack, webhook, iterm2, web)
Implement all 5 notification and terminal UI plugins for the agent
orchestrator, replacing stub files with full implementations of the
Notifier and Terminal interfaces from @agent-orchestrator/core.
Notifier plugins:
- notifier-desktop: OS notifications via osascript (macOS) / notify-send
(Linux) with priority-based sound (urgent=sound, others=silent)
- notifier-slack: Slack Incoming Webhooks with Block Kit formatting,
action buttons, PR links, CI status context blocks
- notifier-webhook: Generic HTTP POST with JSON payloads, configurable
headers, and retry with backoff (default 2 retries)
Terminal plugins:
- terminal-iterm2: AppleScript-based iTerm2 tab management — detects
existing tabs by profile name, creates/reuses tabs (ported from
scripts/open-iterm-tab reference implementation)
- terminal-web: Web terminal session tracking for the dashboard's
xterm.js frontend, providing URL generation and open-state tracking
All plugins follow the PluginModule pattern (manifest + create export)
and pass typecheck cleanly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* test: add comprehensive vitest suites for all notifier and terminal plugins
Add 96 unit tests across 5 plugin packages covering:
- notifier-desktop: priority-based sound mapping, osascript/notify-send
command generation, platform detection (macOS/Linux/unsupported),
title formatting, error propagation
- notifier-slack: Block Kit message structure (header/section/context/
divider blocks), priority emoji mapping, PR link and CI status
rendering, action button generation (URL and callback variants),
channel routing, post method
- notifier-webhook: JSON payload serialization, custom headers, retry
logic (success after retry, exhausted retries, zero retries, network
errors), timestamp ISO serialization
- terminal-iterm2: AppleScript command generation, tab reuse via profile
name detection, new tab creation with tmux attach, runtimeHandle
preference over session ID, batch openAll with delays, error fallback
- terminal-web: session open tracking, dashboard URL configuration,
openAll batch registration, independent state per instance
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address PR review feedback — security, retry logic, side effects
- Add AppleScript injection escaping in terminal-iterm2 and notifier-desktop
- Webhook: only retry on 429/5xx (not 4xx), add exponential backoff
- terminal-iterm2: fix isSessionOpen selecting tab (side effect), remove dead openNewWindow
- notifier-slack: type-guard event.data access, add URL validation
- notifier-webhook: add URL validation, remove unused config interfaces
- Update all tests to cover new behaviors (108 tests passing)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: codex review — shell injection, iTerm2 name lookup, notify-send args, retry validation
- terminal-iterm2: shell-escape session names in tmux command (single-quote wrapping)
- terminal-iterm2: use `name of aSession` instead of `profile name` for tab lookup
- notifier-desktop: fix notify-send arg order (options before title/body)
- notifier-webhook: clamp retries/retryDelayMs to safe values (non-negative, finite)
- notifier-slack: sanitize action_id to [a-z0-9_] characters only
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: lint/format compliance after rebase on main
- Fix ESLint errors: remove unused WebTerminalConfig, ignore next-env.d.ts
- Run prettier on all files for consistent formatting
- Update pnpm-lock.yaml with new lint dependencies
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: codex review round 2 — printf injection, platform guard, config validation
- terminal-iterm2: use shell-escaped name in printf title (not just tmux target)
- terminal-iterm2: add platform guard — no-op with warning on non-macOS
- notifier-webhook: validate customHeaders are string:string before spreading
- notifier-desktop: validate sound config as boolean (reject string "false")
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: update lockfile after rebase on main
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add notifier-composio plugin and integration tests for all plugins
Add a new notifier-composio plugin as the recommended notification
transport using Composio's unified API for Slack, Discord, and Gmail.
Create comprehensive integration tests (79 tests across 6 files) for
all notifier and terminal plugins, mocking only I/O boundaries.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address PR review feedback — security, retry logic, side effects
- Shell injection fix: double-escape (shell + AppleScript) in iTerm2 printf/tmux
- iTerm2 no-window crash: create window if none exists
- Composio graceful degradation: warn instead of throw when SDK missing
- Composio emailTo validation: require emailTo when defaultApp is gmail
- AbortSignal.timeout() replaces manual AbortController + {once: true} listener
- Discord channelName fallback for channel_id
- Slack empty action_id fallback
- Dashboard port: terminal-web default changed from 9847 to 3000
- escapeAppleScript deduplicated into core/utils.ts
- Consistent vitest version (^3.0.0) for composio plugin
- Remove duplicate eslint ignore entry
- Integration tests updated for shared event-factory helper
- Unit tests updated for new validation and escaping behavior
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: update lockfile after rebase on main
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: prevent double ao_ prefix in Slack action_id fallback
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: decouple Linux urgency from sound config, deduplicate validateUrl
- Linux --urgency=critical now driven by event priority, not sound config
- Moved validateUrl to core/utils.ts, imported by slack and webhook plugins
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: handle ESM ERR_MODULE_NOT_FOUND for composio-core detection
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: codex review round 2 — printf injection, platform guard, config validation
- Slack action_id: append index for uniqueness (two "Retry" buttons no
longer collide)
- Composio Gmail subject: extract GMAIL_SUBJECT constant so notify() and
post() use the same value
- Agent integration tests: require API key env vars before running to
prevent CI timeout when secrets are not configured
- Update lockfile after rebase on main
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: parallelize integration tests and increase CI timeout
- Remove singleFork: true from vitest config — all integration test
files use unique session prefixes so they're safe to run concurrently
- Increase CI timeout from 15 to 20 minutes as safety margin for agent
tests that make real API calls
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: prevent unhandled promise rejection after timeout in composio notifier
Attach a no-op .catch() to the executeAction promise so that if the
timeout fires first and the action later rejects, it doesn't trigger
an unhandledRejection event.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
* feat: implement agent plugins (claude-code, codex, aider)
Implement the Agent interface from @agent-orchestrator/core for three
AI coding tools, enabling the orchestrator to launch, detect activity,
and introspect agent sessions.
Claude Code plugin (full implementation):
- JSONL session introspection (summary, session ID, cost, last message type)
- Activity detection via terminal output pattern matching
- Process detection via tmux pane TTY → ps lookup chain
- Launch command generation with permissions/model/prompt support
Codex and Aider plugins (functional stubs):
- Launch command generation with tool-specific flags
- Process detection via TTY lookup
- Stub introspection (no JSONL support yet)
Closes INT-1329
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address PR review — shell escaping, activity detection, tests
- Replace JSON.stringify with POSIX shell escaping (single quotes) for
all agent launch commands to prevent command injection
- Remove `unset CLAUDECODE &&` shell syntax from getLaunchCommand (breaks
execFile); moved to getEnvironment as empty string
- Fix detectActivity returning "exited" for empty pane when process is
confirmed alive — now returns "idle"
- Tighten BLOCKED_PATTERNS to specific error framing (^Error:, ENOENT:,
APIError:, etc.) to avoid false positives on code output
- Check blocked patterns before input patterns to prevent "EACCES:
permission denied" matching INPUT_PATTERNS on the word "permission"
- Remove pgrep fallback in findClaudeProcess (could match wrong session)
- Fix extractCost double-counting: prefer costUSD, fall back to
estimatedCostUsd only when costUSD is absent
- Trim ACTIVE_PATTERNS to stable indicators only
- Add comprehensive vitest tests (127 tests across 3 plugins)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address Codex review + lint fixes
- Use `ps -eo pid,tty,args` instead of `comm` for process detection so
CLI wrappers (node, python) are correctly identified
- Coerce PID from handle.data with Number() + isFinite() to handle
string-serialized PIDs
- Fix token double-counting: prefer `usage` object, only fall back to
flat `inputTokens`/`outputTokens` when `usage` is absent
- Add `--` before Codex prompt to prevent prompts starting with `-`
from being parsed as CLI flags
- Check blocked patterns before input patterns so "EACCES: permission
denied" isn't misclassified as waiting_input
- Fix lint: merge duplicate imports, strict equality, remove non-null
assertions
- Format with prettier
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: multi-pane tmux, EPERM handling, strict process matching
Codex review iteration 1 fixes:
- Iterate all tmux pane TTYs instead of only the first — prevents
false "exited" in multi-pane sessions
- Handle EPERM from process.kill(pid, 0) as "process exists" — EPERM
means the process is alive but we lack permission to signal it
- Use word-boundary regex for process name matching — prevents false
positives on similar names like "claude-code" or paths containing
the substring
- Add tests for EPERM handling, multi-pane detection, and strict
name matching (134 tests total)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: exclude next-env.d.ts from eslint
Auto-generated Next.js type file triggers triple-slash-reference rule.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address Bugbot review — JSON.parse guard, detectActivity pattern priority
- Guard parseJsonlFile against non-object JSON values (null, arrays,
primitives) that would cause TypeError in downstream extractors
- Split null output (non-tmux → "active") from empty output (tmux → "idle")
so non-tmux runtimes don't falsely report idle
- Reorder pattern matching: idle before blocked so stale errors in the
tmux scrollback don't mask an idle prompt
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: extract shellEscape to @agent-orchestrator/core
Move the duplicated POSIX shell-escaping utility from all three agent
plugins into packages/core/src/utils.ts and re-export from the package
entry point. Plugins now import { shellEscape } from the shared package.
Addresses Bugbot review comment on PR #5.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: add execFileAsync timeouts, use satisfies for plugin exports
- Add { timeout: 30_000 } to all execFileAsync calls (tmux, ps) per
CLAUDE.md convention to prevent hanging on stuck processes
- Replace `const plugin: PluginModule<Agent>` with inline
`export default { ... } satisfies PluginModule<Agent>` to preserve
narrow literal types on manifest fields
Addresses Bugbot review comments on PR #5.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* refactor: rename introspect → getSessionInfo, detectActivity uses JSONL
- Rename Agent.introspect() → getSessionInfo(), AgentIntrospection → AgentSessionInfo
- Claude Code detectActivity: replace tmux screen-scraping with JSONL-based
detection (file mtime + last entry type)
- Remove terminal pattern constants (ACTIVE_PATTERNS, IDLE_PATTERNS, etc.)
- Update all tests: JSONL-based activity tests, renamed introspect blocks
- waiting_input/blocked states deferred to hooks-based implementation (#16)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: add OpenCode agent plugin, integration tests, and CI workflow
- Add agent-opencode plugin with getLaunchCommand, detectActivity,
isProcessRunning, getSessionInfo (27 unit tests)
- Add integration test suite for all 4 agents (claude-code, codex,
aider, opencode) using real binaries in tmux sessions
- Add GitHub Actions CI workflow for integration tests
- Add test:integration script to root package.json
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: update core test mocks — introspect → getSessionInfo + isProcessing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* fix: address Bugbot review — Windows paths, bytesRead, hardcoded paths
- toClaudeProjectPath: handle Windows backslashes and drive letters
- readLastJsonlEntry: use bytesRead to avoid null character corruption
- Remove duplicate next-env.d.ts eslint ignore entry
- Remove hardcoded developer-specific binary paths from integration tests
- Add isProcessing limitation comments with issue references (#17-19)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
prateek