Open SWE
An open-source, asynchronous coding agent built on Deep Agents by LangChain that turns a code-change task into a pull request, working in an isolated environment from issue to review.
Open SWE is an open-source, asynchronous coding agent built on Deep Agents by LangChain. Given a code-change task from a dashboard, GitHub, Slack or Linear, or run on a schedule, it works in an isolated environment to understand the codebase, make changes, validate them, and deliver a pull request.
Beyond generating code, it can review pull requests, learn a repository's review style from historical feedback, monitor CI, and respond to feedback; it also supports read-only PR chat for investigating a change without modifying it. Each coding thread is bound to its own persistent sandbox, so the agent can continue prior work when a developer replies, and independent threads run in parallel.
It is open source, deployable on a team's own infrastructure, and designed to be adapted to a team's repositories, tools, policies and workflows. It is written in Python under the MIT license.