Open source platform enabling humans + AI to collaborate for knowledge work β An intelligent GitHub planning system with configurable AI agents that learn from your team's patterns and coordinate tasks across humans and specialized AI assistants.
Autonomy transforms GitHub into an intelligent collaboration platform where AI agents work alongside human team members to plan, prioritize, and coordinate work. The platform supports configurable AI agents - from core planning agents to specialized domain experts - all configurable through simple YAML files. Instead of replacing human judgment, it augments team decision-making with memory-driven insights and automated task coordination.
# Global CLI installation
pipx install autonomy
# Initialize with your GitHub repository
autonomy init --repo your-org/your-repogit clone https://github.com/mehulbhardwaj/autonomy.git
cd autonomy
pip install -e .[dev]
pre-commit install
pytest- Agent Orchestration: Configurable AI agents for different roles (PM, SDE, QA)
- LLM Integration: OpenRouter for unified access to 400+ models
- Memory System: Repository-scoped memory for intelligent context and relationship tracking
- GitHub Integration: GraphQL + REST for comprehensive GitHub automation
- Team Communication: Slack integration for notifications and collaboration
- CLI Interface: Planning, ranking, assignment, status, and more
- GitHub Integration: Issues, Projects v2, authentication, board bootstrap, hierarchy management
- Slack Integration: Basic bot, slash commands, notifications
- Task Management: Priority ranking, manual override (pin/unpin), breakdown, dependencies
- Hierarchy Sync: Auto-create parents with Tasklists API
- Undo System: Shadow branch PRs and diff hash restore
- Metrics & Notifications: Daily metrics reports and Slack alerts
- Security & Storage: OS-native secret storage, basic audit logging
- Documentation: Comprehensive user and developer docs
- Installation Guide - Get started with Autonomy
- User Guide - Learn how to use Autonomy CLI and features
- Configuration - All configuration options and settings
- Technical Architecture - System design and implementation details
- Product Requirements - Product vision, roadmap, and development workflow
- Implementation Timeline - Week-by-week delivery plan
- Testing Strategy - Testing approach and coverage strategy
This project maintains focused documentation with clear separation of concerns:
- User-facing: Installation, usage, and configuration guides
- Developer-facing: Architecture, requirements, implementation, and testing
- Cross-references: Each document references related documentation where appropriate
The examples/ directory contains configuration templates and examples:
agent.yml- Example AI agent configurationboard_cache.json- GitHub Projects field cache example
Basic Setup:
# Install and authenticate
pipx install autonomy
autonomy auth login
# Initialize repository
autonomy init --repo my-org/my-repo
# Get next task
autonomy nextSlack Integration:
# Setup Slack bot
autonomy auth slack install
# Use Slack commands
/autonomy next
/autonomy update 123 --done
/autonomy pin 456Board Management:
# Initialize GitHub Projects board
autonomy board init
# Rank items by priority
autonomy board rank
# Reorder items
autonomy board reorder
# Sync hierarchy
autonomy hierarchy-syncUndo Operations:
# Undo the last change
autonomy undo --last
# Specify a custom commit window
autonomy undo <hash> --commit-window 3
# Generate a shadow branch PR for recent changes
autonomy audit shadow-pr --limit 3We welcome contributions from developers interested in AI-human collaboration and intelligent automation!
- Read the docs: Start with ARCHITECTURE.md and REQUIREMENTS.md
- Setup environment: Follow INSTALLATION.md for development setup
- Pick an issue: Check good first issues
- Join discussions: Participate in issue discussions and planning
Stable releases are tagged from the main branch. Testing releases use commits from the testing branch and are published as pre-releases. Development work should occur on feature branches that merge into testing before stabilizing in main.
This project is licensed under the GNU GPLv3 or later - see the LICENSE file for details.
Status: Active open source development. Contributions welcome!
