PatchPilot

Project description

PatchPilot coordinates focused agents, scores the quality of each completed run, and uses reviewer feedback to improve the next attempt. The team built the full workflow during the hackathon, using Guild.ai for the core agent loop and Actian for evaluation, communication, or runtime support. The demo includes a visible activity trace so judges can inspect what each agent did and where human approval was required.

Project links

• GitHub repository
• Project website
• Demo video

Technical architecture

A LangGraph supervisor assigns investigation, reproduction, patching, and verification work to specialized agents, then keeps the best result in shared state.

Setup instructions

Clone the public repository, install dependencies, copy .env.example to .env.local, add the documented API keys, and run the development command in the README.

Lessons learned

The team found that clear handoffs and fast evaluation loops mattered more than adding more agents. Guild.ai handled orchestration while Actian made the result easier to inspect and improve.

Additional screenshots or links

• https://docs.langchain.com/oss/python/langgraph/overview
• https://github.com/langchain-ai/langgraph