Teaching an AI Coding Agent to Review Its Own Work with Code Meridian

Here I am refactoring:

Feedback from evaluation:

Suggestion for improving the tool:

Code Meridian is an open-source context tool for AI coding agents. It indexes a project into a graph so agents can reason about relationships in the codebase instead of only reading isolated files.

https://github.com/Driftya/code-meridian

Learn more:
https://driftya.com/content/how-codemeridian-helps-me-work-on-driftya