# Code Review > Reviews changes for correctness, security, maintainability, and alignment with Brand-Guard Cursor rules (TypeScript, React, clean code, git flow). Use when reviewing pull requests, diffs, staged changes, or when the user asks for a code review. Use when this capability is needed. - Skill: `tomevault-io/code-review-158` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/code-review-158` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/code-review-158/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Web & Frontend - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/code-review-158 --- # Code review (Brand-Guard) ## Before reviewing 1. Read the diff or files in scope; prefer whole-function context over single hunks when assessing logic. 2. For stack-specific depth, open [reference.md](reference.md) and load only the listed rule files that match the change. ## Review checklist - **Correctness**: Logic matches intent; edge cases and error paths handled; no regressions for callers. - **Security**: Input validation, XSS/authz/session handling, secrets not logged or committed, safe handling of URLs and HTML. - **TypeScript**: No unnecessary `any`; sound narrowing; explicit public APIs where the project expects them. - **React**: Hooks rules, effect deps and cleanup, keys in lists, avoid avoidable re-renders. - **Maintainability**: Single responsibility, meaningful names, DRY without premature abstraction; matches surrounding patterns. - **Tests**: Critical paths and regressions covered where the project tests similar code. - **Scope**: Change stays focused; unrelated code and drive-by refactors flagged only when risky. ## Feedback format Use severity labels (no emoji required): - **Critical**: Must fix before merge (bugs, security, data loss, broken contracts). - **Suggestion**: Should fix or discuss (maintainability, clearer types, smaller components). - **Nice to have**: Optional polish. For each finding: short title, why it matters, and a concrete fix or alternative. Use code citations with ```startLine:endLine:path when referencing existing code. ## PR / merge readiness - Call out missing migrations, env changes, or breaking API changes. - If git-flow applies: branch naming and target branch match [reference.md](reference.md) git-flow summary. ## Anti-patterns during review - Do not nitpick pure formatting or whitespace-only churn. - Do not propose large unrelated refactors unless blocking correctness or security. - Verify claims against the code shown; flag uncertainty explicitly. --- > Source: [KAMRONBEK/Brand-Guard](https://github.com/KAMRONBEK/Brand-Guard) — distributed by [TomeVault](https://tomevault.io).