uygnoey Deep Review
Run three parallel subagent reviews on the staged changes and merge their findings into one prioritized summary.
Instructions
- Security review — check for vulnerabilities, injection risks, authentication issues, and sensitive data exposure.
- Performance review — check for N+1 queries, unnecessary iterations, memory leaks, and blocking operations.
- Style review — check for consistency with project patterns documented in
/docs/style-guide.md.
Run all three in parallel (not sequentially). Each reviewer should use read-only tools only.
Output format
Synthesize findings into a single summary with priority-ranked issues. Each issue must include:
- File and line number (
path/to/file.ext:123) - Category (security / performance / style)
- Severity (high / medium / low)
- Recommended fix
If a reviewer finds nothing, it must say so explicitly rather than inventing issues.
Examples
Invocation in a Claude Code session:
/deep-review
Expected synthesized output shape:
High severity
- src/api/payments.ts:142 (security) — User input flows into a raw SQL string. Parameterize.
- src/db/pool.ts:37 (performance) — N+1 inside a loop over orders. Batch-fetch once.
Medium severity
- src/components/Cart.tsx:88 (style) — Diverges from the list-rendering pattern in /docs/style-guide.md.
No findings
- Style reviewer reports no additional style deviations.
Source
Distilled verbatim from How and when to use subagents in Claude Code (published 2026-04-07).