Gavel MCP server

Gavel is a code-quality checker for big Bazel projects. It looks at the code you just changed and tells you if it broke any of your team's rules for lint, test coverage, or code structure. It is handy if you work in a large codebase and want your AI assistant to catch problems while it writes, not later in the build.

What this MCP server does

You ask your AI to check some code or a change you just made. The AI calls this helper, which runs Gavel's checks on your Bazel project. Gavel looks only at the files and targets that changed, compares them to a saved baseline, and decides if anything new broke the rules. It sends back a verdict with the details, and your AI shows you the result in the chat.

Click to zoom

What you can do with it

• Run the full quality gate and get a pass or fail verdict
• Lint a single file and see the findings
• List new, fixed, and existing code issues
• Check test coverage numbers and the trend
• Check for architecture rule violations
• See only the findings caused by your latest change

Try asking your AI

• “Judge my current changes and tell me if the quality gate passes”
• “Lint the file I just edited and show me any new problems”
• “What is the coverage for the payments project right now”
• “Are there any new architecture violations in my change”

What it gives back to you

You get a verdict back: PASS or FAIL, plus a short reason. Along with it come lists of findings marked as new, fixed, or existing, coverage percentages with the change since last time, and any architecture problems. Your AI shows this as a summary in the chat, and you can ask follow-up questions about any item.

Before you start

Who it's for

Developers and teams working in large Bazel monorepos who want their AI coding assistant to respect the project's quality rules.