MergeFlow AI

MergeFlow AI is an AI-powered post-merge automation agent for GitHub pull requests.

It listens to GitHub webhook events, runs a Celery-based automation pipeline, and helps teams handle repetitive post-merge work such as pull request review, diff classification, issue updates, environment variable detection, and generated documentation tasks.

What It Does

• Listens for GitHub pull request webhook events
• Runs pre-merge AI self-review when a PR is opened
• Runs post-merge automation when a PR is merged
• Classifies changed code as API, frontend, database, infrastructure, or mixed
• Moves linked GitHub issues to done by closing completed issues
• Detects new environment variables and updates .env.example
• Provides foundation for changelog, Swagger, Notion, SendGrid, RAG, and dependency graph features

Tech Stack

• Python
• FastAPI
• Celery
• Redis
• Docker / Docker Compose
• GitHub API
• Gemini API
• ChromaDB-ready RAG modules

Project Structure

MergeFlow-AI/
├── backend/
│   ├── main.py
│   ├── worker.py
│   ├── classifier/
│   ├── features/
│   ├── mcp_servers/
│   ├── rag/
│   └── websocket/
├── frontend/
├── docker-compose.yml
├── Dockerfile
├── requirements.txt
├── .env.example
└── README.md