SWE-agent

Open-source autonomous software engineering agent

Code & DevelopmentFree

SWE-agent is an open-source AI agent system from Princeton NLP that can autonomously fix bugs in real GitHub repositories. It provides a computer interface for LLMs to interact with code, run tests, and submit patches. Researchers and developers use SWE-agent for studying and building autonomous coding agents.

Key Features

• ✓Autonomous bug fixing
• ✓GitHub issue resolution
• ✓ACIformats for LLM interaction
• ✓Multiple LLM backend support
• ✓Research benchmark evaluation