Open SWE
Open-source coding agent that takes engineering tasks and delivers pull requests, meant to be self-hosted.
- Stars
- 11.2k
- Forks
- 1.3k
- Licence
- MIT
- Language
- Python
- Last update
- 4 days ago
- Type
- AI model / agent
What it is
Open SWE is an open-source coding agent from LangChain that investigates a codebase, makes changes in an isolated sandbox, validates them, and opens a pull request. It can also review pull requests, monitor CI, and respond to feedback. It is built on Deep Agents and LangGraph and is designed to be deployed in your own infrastructure.
- Start a coding task from a GitHub issue, Slack, or the dashboard and get a pull request back
- Run on-demand or opt-in automatic reviews that publish findings to GitHub
- Ask read-only questions about a PR through PR chat without changing code
- Schedule recurring tasks or monitor opted-in PRs with /baby-sit
Great for
- Engineering teams that want a self-hosted coding agent they can inspect and adapt
- Developers who already run LangChain or LangGraph and want a reference implementation
- Teams that want an agent to handle routine PR work such as reviews and CI follow-ups
- Organizations that need control over sandbox providers, models, and GitHub access
Not ideal for
- Anyone wanting a ready-to-use hosted product: the README says deploying as-is requires Managed Deep Agents, otherwise you must fork and adapt the code to your own runtime
- Teams that expect stable releases: the README says it is under active development with expected breaking changes
- People hoping to contribute issues or patches: the project is not accepting issues or external contributions at this time
- Anyone who needs strong isolation for local execution: the README says local execution does not provide cloud sandbox isolation
How to get it
- 1
Clone and build from source
Clone the repository
git clone https://github.com/langchain-ai/open-swe - 2
Run it
- Read the installation guide at docs/INSTALLATION.md
- Configure your deployment, including the LangSmith sandbox or another provider
- Connect GitHub access with the scope you need
- Start a task from the dashboard, a GitHub issue, or Slack
- 3
Get the most out of it
- Use the customization guide to choose models, reasoning effort, skills, and sandbox providers
- Scope MCP connections per instance, workspace, or person using the customization guide
- Configure workspace repository bindings and settings in the workspaces reference
- Set up Slack human review for pull requests using the human-review docs
- Review the generated OpenAPI schema in swagger.json for API integration
Downloads
Strengths
- Covers the full loop from investigation to pull request, review, and CI follow-up
- Open source under MIT, and deployable in your own infrastructure
- Lets you choose models, sandbox providers, integrations, and approval policies
Trade-offs
- Under active development with breaking changes and no stability guarantee
- Deployment as-is depends on Managed Deep Agents; otherwise you must adapt it yourself
- Sandboxes have powerful tools and may have network access, so credentials must be tightly scoped
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| open-swe (this) | 11.2k | AI model / agent | advanced | 4 days ago | Engineering teams that want a self-hosted coding agent they can inspect and adapt |
| Prompt-Engineering-Guide | 79.8k | Learning resource | beginner | 8 months ago | People who are new to prompting and want structured lessons with examples |
| langchain | 149k | Framework | intermediate | 4 days ago | Python developers building LLM-powered applications |
| SuperAGI | 17.7k | AI model / agent | advanced | 1.8 years ago | Python developers experimenting with autonomous AI agents |
| OpenHands | 92.5k | App | intermediate | 4 days ago | Developers who want coding agents running on a server while their laptop is off |
| crewAI | 60.4k | Library | intermediate | 4 days ago | Python developers prototyping multi-agent workflows |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.