review-full

This is a COMPREHENSIVE multi-agent code review. For quick anti-pattern scanning (seconds, not minutes), use review-diff instead. ## Step 0: Detect scope and project - If `$ARGUMENTS` is empty: review uncommitted changes (staged + unstaged via `git diff` + `git diff --cached`) - If `$ARGUMENTS` is a file path: review that file only - If `$ARGUMENTS` is a branch or range: review diff against that ref - If `$ARGUMENTS` is "security": run security-only review (skip to step 3) - Detect project type from cwd/CLAUDE.md and the project manifest (any language/framework), or fall back to inspecting the file structure ## Step 1: Spawn review agents in parallel Launch up to 4 agents based on what the changes touch: | Changes Touch | Agent to Spawn | Focus | |--------------|----------------|-------| | Any code | `code-reviewer` | Quality, patterns, naming, DRY, error handling, consistency | | API endpoints, auth, user input | `security-reviewer` | OWASP Top 10, injection, auth gaps, secrets, CORS | | Database queries, ORM models, migrations | `db-analyst` | N+1, undefined vs null, missing models, query performance | | New modules, moved files, cross-layer imports, large refactors | `architecture-reviewer` | Dependency direction, circular deps, god files, dead code, modularity | If changes are small (<50 lines), run code-reviewer only. If security argument, run security-reviewer only. Add `architecture-reviewer` when the change reshapes structure (new module boundaries, files moved ...