What problem does it solve? Diff-only code reviews miss bugs that live outside the changed lines: callers whose contracts no longer match, duplicate implementations needing the same fix, and downstream modules that consume changed values. This Skill forces a structured deep review that reads surrounding code before judging a PR. ## Core Features & Use Cases - Four-lens evaluation: Judges specification compliance, code correctness (bugs, security, performance), code quality (naming, complexity, duplication), and overall design consistency as separate dimensions. - Mandatory context gathering: Enumerates callers, similar implementations, related tests, and downstream consumers of changed symbols via grep before any verdict is issued. - Finding verification and parallel execution: Important findings must be backed by actually-read files, and large PRs can be fanned out to parallel review agents per lens. - Use Case: Before merging a multi-file refactor, ask for a deep review to confirm the changed function's new contract matches every caller and that no sibling files still carry the old buggy pattern. ## Quick Start Ask the AI to deeply review the current branch diff against main, including surrounding code, callers, and similar implementations.
