llms.txt

# Pilot

> Autonomous AI development agent that converts tickets into pull requests. Label a GitHub issue with "pilot" and Pilot handles implementation, testing, and PR creation.

## Quick Integration

**Trigger Pilot**: Create a GitHub issue with the `pilot` label.

**Issue Format**:
```
Title: Short description of the task
Body:
- What needs to be done
- Acceptance criteria
- Any constraints or preferences
```

**What Happens**:
1. Pilot claims the issue (adds `pilot/in-progress` label)
2. Creates branch `pilot/GH-{number}`
3. Analyzes codebase and plans implementation
4. Writes code using Claude Code
5. Runs tests, lint, build validation
6. Opens PR linked to issue (adds `pilot/done` label)
7. Sends notification via Telegram/Slack

**Supported Inputs**:
- GitHub Issues (polling, `pilot` label)
- Linear (webhooks)
- Jira (webhooks)
- Asana (webhooks)
- Telegram Bot (direct messages)

**Best For**: Bug fixes, small features, refactoring, tests, documentation, dependency updates.

**Not Ideal For**: Large architectural changes, security-critical code, tasks requiring human judgment.

## Documentation

- [README](https://github.com/qf-studio/pilot#readme): Full documentation, installation, configuration
- [Deployment Guide](https://github.com/qf-studio/pilot/blob/main/docs/DEPLOYMENT.md): Cloud deployment options
- [CLI Reference](https://github.com/qf-studio/pilot#cli-reference): All commands and flags

## Optional

- [Configuration](https://github.com/qf-studio/pilot#configuration): YAML config reference
- [Autopilot Modes](https://github.com/qf-studio/pilot#autopilot-modes): dev/stage/prod autonomy levels
- [Telegram Integration](https://github.com/qf-studio/pilot#telegram-integration): Chat, voice, and image support