What it does
Gito is an MIT-licensed tool that asks a language model to review code changes and report issues such as security vulnerabilities, bugs and maintainability concerns. It can review a GitHub pull request from a GitHub Actions workflow, post the result as a comment and react to maintainer comments. GitLab support is marked beta, Bitbucket is listed as planned, and local review from the command line is supported. You choose the model. The page lists OpenAI-compatible APIs (including Mistral, xAI, Azure, Amazon Bedrock, OpenRouter and Fireworks), the Anthropic and Google APIs, local services such as Ollama, vLLM, llama.cpp, SGLang and LM Studio, and CLI coding agents such as Claude Code and Gemini CLI. It says code goes straight from your CI runner or machine to the configured provider, with no intermediary server and nothing stored or logged by Gito. You can install it with pip or uvx, or with a Windows standalone installer. The gito setup command runs a wizard for the LLM connection, gito review checks the current branch against the main branch, and gito deploy can generate CI workflow files. Review prompts, output templates and mention triggers are set in a .gito/config.toml file in the repository, and Jira and Linear integrations are listed.
Capabilities
- Reviews GitHub pull requests from GitHub Actions and posts the result as a comment
- Reviews the current branch locally from the command line
- Works with OpenAI-compatible, Anthropic and Google APIs and local model servers
- GitLab CI support (beta) plus Jira and Linear integrations
- Per-repository config for prompts, report templates and triggers
When to use it
- You want automated review comments on pull requests without a middleman service seeing your code
- You already use a particular LLM provider or local model and want a reviewer that works with it
- You want to review a branch locally before opening a pull request
- You want review rules kept in version control with the repository
