codebase-audit

A whole-repository codebase audit at HEAD identifies cross-cutting specification, logic, and product-security defects that diff-scoped reviews miss. It runs parallel read-only agent reviews across the accumulated repository to detect specification drift, wiring failures, and redundancy.

The best way to consolidate duplicate code findings is to perform class-level finding management during a codebase audit. This sweeps each defect pattern across the repository, deduplicates occurrences, batches documentation drift, and creates controlled draft PBIs for remediation.

A codebase audit checks spec conformance before integration testing by running a thin integration-entry re-check. This blocks testing when unresolved Critical or High findings remain, verifying the repository is ready by reviewing enabled specifications, requirements, and test suites.

Yes, you can use a codebase audit for non-blocking cross-review ceremonies at the end of a Development Sprint. It inspects the repository at HEAD, synthesizes findings, and routes remediation PBIs according to severity without halting ongoing development.

Static analysis at HEAD detects product security gaps by auditing the accumulated repository across four parallel axes, including product security. It identifies whole-repository vulnerabilities and stale documentation before integration testing begins.