SWE-agent lets LLMs autonomously fix GitHub issues and find security vulnerabilities. 18.9K+ stars. State-of-the-art on SWE-bench. MIT.
这里应该同时让用户和 Agent 知道第一步该复制什么、安装什么、落到哪里。
# Install
pip install sweagent
# Fix a GitHub issue
sweagent run --agent.model.name=claude-sonnet-4-20250514 \
--problem_statement.github_url=https://github.com/owner/repo/issues/123
# Or run on SWE-bench
sweagent run --problem_statement.swe_bench_id=django__django-11905SWE-agent enables language models like GPT-4o or Claude Sonnet to autonomously fix GitHub issues, identify cybersecurity vulnerabilities, or tackle custom coding challenges. With 18,900+ GitHub stars and MIT license from Princeton and Stanford, it achieves state-of-the-art performance on SWE-bench among open-source tools. Features flexible agent design, single YAML configuration, and a research-friendly hackable codebase.
Best for: Teams wanting AI to automatically resolve GitHub issues and bugs Works with: Claude Code, OpenAI Codex, Cursor, Gemini CLI, Windsurf
Q: What is SWE-agent? A: Autonomous coding agent with 18.9K+ stars that fixes GitHub issues using LLMs. State-of-the-art on SWE-bench. Princeton/Stanford. MIT.
Q: How do I install SWE-agent?
A: pip install sweagent. Then sweagent run --problem_statement.github_url=
SWE-agent/SWE-agent — 18,900+ GitHub stars
TaskWeaver is a Microsoft code-first agent framework for data analytics tasks. 6.1K+ stars. Planning, stateful execution, DataFrames, plugins. MIT.
AIChat is a CLI for 20+ LLM providers with shell assistant, RAG, agents, and function calling. 9.7K+ stars. MIT/Apache 2.0.
Khoj is a personal AI app for chat, search, and knowledge management. 33.8K+ stars. Multi-LLM, docs, Obsidian, WhatsApp, custom agents. AGPL-3.0.