SWE-agent
Product & DevOpen SourceAI Coding Assistant
Autonomous AI agent that fixes GitHub issues and implements code tasks.
About
SWE-agent is an autonomous coding agent that enables language models like GPT-4o and Claude Sonnet to independently identify and fix issues in real GitHub repositories. It automatically uses tools to write, analyze, and improve code across any programming language. Built by Cognition Labs (makers of Devin), SWE-agent represents a major advancement in AI-assisted software engineering, achieving 40.2% pass@1 on SWE-bench Verified with its specialized 32B model.
Problem it solves
Enabling AI language models to autonomously identify, implement, and resolve software engineering tasks without human intervention.
Best for
Engineering teams, DevOps teams, and software development organizations seeking to automate code review and issue resolution at scale.