agent-orchestrator/packages/plugins/runtime-tmux
Harsh Batheja 4e2144d99e
feat: OpenCode session lifecycle and CLI controls (#315)
* feat: refine OpenCode session reuse strategy and cleanup

* fix: harden OpenCode session selection and lint errors

* refactor: centralize OpenCode reuse resolution flow

* fix: return 404 for missing session in message route

* feat: replace force remap with terminal reload control

* fix: protect project path from session kill cleanup

* fix: preserve fullscreen alignment without reload action

* fix: harden OpenCode session id handling and title reuse selection

* fix: show OpenCode reload control and remap before restart

* fix: preserve title-only OpenCode reuse with fallback mapping persistence

* fix: resolve remaining PR315 Bugbot findings

* fix: keep OpenCode discovery title-based without timestamp sorting

* fix: avoid enrichment race fallout in session listing

* fix: guard OpenCode discovery parse with array check

* fix: stabilize Linear comment integration check

* fix: harden OpenCode discovery and prompt option flow

* ux: clarify OpenCode terminal restart action

* fix: remap OpenCode session fresh on each restart

* fix: validate remap session ids before reuse

* fix: clean archived metadata only after purge

* docs: align OpenCode remap selection with title-based behavior

* fix: harden opencode cleanup and ignore local sisyphus state

* test: add timeout cleanup coverage for session enrichment

* fix: harden Linear integration helper against transient non-JSON errors

* fix: make linear integration assertions resilient to eventual consistency

* fix: remove unused fs import after rebase

* fix: address remaining Bugbot blockers for opencode session handling

* fix: avoid stale metadata overwrite during restore post-launch

* fix: forward subagent in orchestrator flows and defer reuse lookup

* fix: apply configured subagent fallback for session spawn

* fix: scope archived cleanup by project and delay archive restore write

* fix: normalize orchestrator strategy aliases in start display logic

* fix: centralize orchestrator strategy normalization in core

* fix: derive orchestrator reuse display from spawn result

* fix: keep cleanup results consistent across project-id collisions

* fix: namespace cleanup results when session IDs collide

* fix: harden GitHub issue stateReason fallback

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: avoid false failing CI state mapping

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: add tmux command timeouts

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: bound session API enrichment latency

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: repair scm-github merge resolution

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: repair lifecycle-manager test merge

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: delay archive metadata recreation until restore passes

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* test: restore claim-pr session mocks

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: address review findings for OpenCode lifecycle PR

- Fix stripControlChars to preserve newlines for reload commands
- Add SessionNotFoundError and use instanceof checks in API routes
- Document orchestratorSessionStrategy in YAML example
- Validate existingSessionId with asValidOpenCodeSessionId()
- Extract inline Node script to buildSessionLookupScript helper
- Create OpenCodeSessionManager interface for remap capability
- Create OpenCodeAgentConfig type for agent-specific config
- Change default orchestratorSessionStrategy from delete to reuse

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix: guard reused session display without metadata

* chore: add agent config files to .gitignore

Agent configuration files (CLAUDE.md, AGENTS.md, IMPROVEMENTS.md, etc.) are personal and project-specific. They should not be committed to the repository.

Changes:
- Remove CLAUDE.md from git tracking
- Add agent config files to .gitignore
- Create .gitignore-template for reference

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* chore: update gitignore for agent config folder structure

Reorganized agent configuration files:
- CLAUDE.md and AGENTS.md stay in root (agents read them there)
- Tracking files move to .opencode/ (IMPROVEMENTS.md, etc.)
- Optional Claude files in .claude/

Updated .gitignore to ignore folders instead of individual files.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* fix: tighten opencode remap and session discovery safeguards

* fix: address Bugbot findings in session manager

* fix: scope opencode discovery to opencode sessions

* fix: restore concurrent listing and strict permission literals

* fix: throw SessionNotFoundError, parallelize list enrichment, fix permissions type

- Session manager now throws SessionNotFoundError instead of plain Error
  for missing sessions, so web API routes correctly return 404 (not 500)
- Parallelize session enrichment in list() — was sequential, causing O(N)
  latency for N sessions with subprocess enrichment
- Fix AgentLaunchConfig.permissions type to accept legacy "skip" value
  (AgentPermissionInput instead of AgentPermissionMode)
- Add happy-path and validation tests for /api/sessions/:id/message route
- Update all test mocks to use SessionNotFoundError

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: route ao send through session manager

* feat: add purge option to orchestrator stop

* fix: register opencode agent in web services

* test: align web API missing-session coverage

* fix: match notifier config by plugin name

* refactor: dedupe session lookup and tmux buffer send flow

* test: update send lifecycle wait expectation

* fix: harden send routing and cleanup purge controls

---------

Co-authored-by: Harsh <harsh@Ubuntu-24-Forrest.lan>
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: Prateek <karnalprateek@gmail.com>
2026-03-08 09:55:44 +05:30
..
src feat: OpenCode session lifecycle and CLI controls (#315) 2026-03-08 09:55:44 +05:30
README.md feat: seamless onboarding with enhanced documentation (#66) 2026-02-16 22:22:13 +05:30
package.json feat: publish to npm under @composio scope (#32) 2026-02-15 04:28:57 +05:30
tsconfig.json feat: scaffold TypeScript monorepo with all plugin interfaces 2026-02-13 17:02:42 +05:30

README.md

@agent-orchestrator/plugin-runtime-tmux

Runtime plugin for executing agent sessions in tmux.

What This Does

Creates isolated tmux sessions for each agent. Each session runs in a separate tmux session with:

  • Working directory set to workspace path
  • Environment variables from config
  • Agent launch command executed automatically

How It Works

Creating a Session

const handle = await runtime.create({
  sessionId: "my-app-3",
  workspacePath: "/Users/dev/.worktrees/my-app/my-app-3",
  launchCommand: "claude -p 'Fix bug in auth module'",
  environment: {
    AO_SESSION_ID: "my-app-3",
    AO_PROJECT_ID: "my-app",
  },
});

What happens:

  1. Validates sessionId (only alphanumeric, dash, underscore allowed)
  2. Creates detached tmux session: tmux new-session -d -s my-app-3 -c /path/to/workspace
  3. Sets environment variables: tmux ... -e KEY=VALUE
  4. Sends launch command: tmux send-keys -t my-app-3 "claude -p '...'" Enter
  5. Returns RuntimeHandle with tmux session name

Sending Messages

await runtime.sendMessage(handle, "Fix the test failure in auth.test.ts");

What happens:

  1. Clears partial input: tmux send-keys -t my-app-3 C-u
  2. For short messages (<200 chars, no newlines): sends directly with -l flag (literal mode)
  3. For long/multiline messages: writes to temp file → tmux load-buffertmux paste-buffer
  4. Waits 300ms (let tmux process the text)
  5. Sends Enter: tmux send-keys -t my-app-3 Enter

Why the complexity?

  • send-keys without -l interprets special strings ("Enter", "Space") as key names
  • Long strings can overflow tmux's command buffer
  • Multiline strings need special handling

Getting Output

const output = await runtime.getOutput(handle, 50); // last 50 lines

Uses tmux capture-pane -t my-app-3 -p -S -50 to capture terminal buffer.

Checking if Alive

const alive = await runtime.isAlive(handle);

Uses tmux has-session -t my-app-3 (exit code 0 = exists, 1 = doesn't exist).

Destroying

await runtime.destroy(handle);

Kills tmux session: tmux kill-session -t my-app-3 (ignores errors if already dead).

Attaching to Sessions

For Terminal plugins (iTerm2, web):

const attachInfo = await runtime.getAttachInfo(handle);
// Returns: { type: "tmux", target: "my-app-3", command: "tmux attach -t my-app-3" }

Security

Session ID validation:

const SAFE_SESSION_ID = /^[a-zA-Z0-9_-]+$/;

Only allows safe characters. Prevents shell injection via session name (used in tmux commands).

Error Handling

  • Session creation fails → cleans up (kills session) before throwing
  • Message send fails → throws (caller should handle)
  • Session already deaddestroy() silently succeeds (idempotent)

Metrics

const metrics = await runtime.getMetrics(handle);
// Returns: { uptimeMs: 123456 }

Tracks uptime (stored in RuntimeHandle.data.createdAt).

Testing

This plugin is tested indirectly via packages/core/src/__tests__/tmux.test.ts (utility functions) and integration tests.

To test manually:

# Start a test session
tmux new-session -d -s test-session -c /tmp
tmux send-keys -t test-session "echo hello" Enter

# Capture output
tmux capture-pane -t test-session -p

# Kill session
tmux kill-session -t test-session

Common Issues

tmux not installed

If tmux is not in PATH, all operations fail. Install via:

  • macOS: brew install tmux
  • Linux: apt-get install tmux or yum install tmux

Session name conflicts

If a session with the same ID already exists, create() fails. The orchestrator should ensure unique session IDs.

Detached sessions persist after orchestrator crashes

tmux sessions keep running even if the orchestrator dies. Use tmux list-sessions to find orphans, tmux kill-session -t <name> to clean up.

Limitations

  • macOS/Linux only — tmux is not available on Windows (use WSL)
  • No Windows native support — use runtime-process instead on Windows
  • Terminal buffer sizegetOutput() limited by tmux buffer size (default 2000 lines)
  • No resource limits — agents can consume unlimited CPU/memory (use docker/k8s runtimes for isolation)

Architecture Notes

Why tmux over raw processes?

  • Sessions persist across orchestrator restarts
  • Easy to attach for debugging: tmux attach -t session-name
  • Terminal emulation (colors, ANSI codes work)
  • Works well with interactive AI tools (Claude Code, Aider)

Why detached mode?

  • Orchestrator doesn't block waiting for agent
  • Multiple agents can run in parallel
  • Humans can attach later without interrupting agent