What it does
Rules such as “needlessly complicated” can appear beside ordinary diagnostics while remaining visibly judgment calls.
Agent tooling
Turns subjective code-quality rules written in JSON into stable command-line and editor diagnostics.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
Rules such as “needlessly complicated” can appear beside ordinary diagnostics while remaining visibly judgment calls.
Use this to flag code that your team finds hard to read. Ask a developer to start with the sample rule and preview which code it would send. Live checks share that code with an outside AI service. Treat warnings as opinions to discuss, not proof that the code is wrong.