[KongchangAI]
Product

GitWarren

A locally run, open-source code review tool that supports connecting to AI agents via the MCP protocol before Git commits to review code changes across four states in the working tree. It provides inline comments and threaded comment management.

Timeline (last 90 days)

Sep 12

GitWarren is a locally running code review application that operates directly on the working tree, enabling a Pull Request-like review experience without needing to push to a remote server.

Unverified50%
Sep 12

GitWarren can review code changes in four states: committed, staged, unstaged, and untracked new files.

Unverified50%
Sep 12

GitWarren supports leaving inline comments on specific lines of code, organized into independent review units in thread form.

Unverified50%
Sep 12

GitWarren connects AI tools via MCP (Model Context Protocol), allowing AI agents to directly read working tree changes and provide feedback in review threads

Unverified50%
Sep 12

GitWarren is based on MCP rather than being bound to a specific model, remaining neutral toward AI vendors, and any coding agent that supports MCP can theoretically be integrated

Unverified50%
Sep 12

GitWarren is an open-source developer tool created by independent developer Michał Wrzosek

Unverified50%
Sep 12

Traditional code review typically occurs after committing and pushing to a remote repository and opening a Pull Request, whereas GitWarren moves the code review point forward to before the commit

Unverified50%
Sep 12

GitWarren does not attempt to replace GitHub's PR workflow, but rather supplements the phase before committing

Unverified50%

All Facts (8)

Source Articles