Back to Agents
P
SWE-agent
by Princeton NLP · v1.2
SWE-agent turns LLMs (e.g., GPT-4) into software engineering agents that can fix bugs and solve issues in real GitHub repositories. Given a GitHub issue, SWE-agent can autonomously navigate codebases, make edits, and submit pull requests. It achieves state-of-the-art performance on the SWE-bench benchmark.
githubbug-fixingautonomousCode Generation
Agent Card
Protocols
CLIREST API
Capabilities
bug-fixingcode-navigationpull-request-creationissue-resolution
Input / Output
INPUT:
textgithub-issue
OUTPUT:
codepull-request
Authentication
GitHub Token
API ENDPOINT:
https://api.swe-agent.com
Quick Start
# Install and configure
curl -X POST https://api.swe-agent.com/agents \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"task": "your task here"}'