io.github.zengxzh/buer

Enable buer in my AI coding agent workflow to monitor repeated rewrites of the same file, recurring patches, or stalled progress; if a loop is detected, stop execution and report the reason and related files.

A loop-risk alert explaining the trigger, listing affected files, and stopping further unproductive agent edits.

Before the agent submits a fix, use buer to check whether the new changes introduce regressions; assess risk using test results, change scope, and past behavior, and flag high-risk submissions.

A regression risk assessment highlighting suspicious changes, affected tests or modules, and whether the submission should be blocked.

Use buer to review changes the agent made to pass tests, detecting removed assertions, weakened conditions, skipped cases, or test edits that hide defects, and produce an audit summary.

An audit report on test tampering that lists suspicious test changes, risk levels, and recommended manual review items.