review

An evidence-based codebase review validates each security problem, architecture smell, and operational risk with concrete proof like file paths and grep outputs. This Skill runs a multi-phase protocol to investigate your codebase and outputs a prioritized findings document with remediation tasks.

Yes, you can scope a code audit by passing an argument to narrow the review. You can target full, security, consistency, or hygiene scopes, allowing faster targeted audits that focus only on specific categories like documentation drift or dead code.

The best way to identify dead code and test gaps is to run a structured multi-category review protocol. This process investigates stale artifacts and test coverage gaps, requiring concrete evidence for every reported issue before writing the results into a prioritized findings file.

A structured code review generates a remediation plan by validating findings with concrete evidence and grouping them by severity. It writes a prioritized findings document to docs/review-YYYY-MM-DD.md, containing executable remediation tasks grouped by critical, high, medium, and low severity.

No, this code review tool requires no external dependencies to check for inconsistencies. It applies an opinionated, multi-phase protocol directly to your codebase to investigate consistency, standards, and deployment risks without needing additional components.

You should not use a full codebase review when you only need to address a specific category like security or consistency. Instead, use the scope control argument to run a targeted audit, saving time by skipping unrelated categories like documentation drift or operational risks.