# Review > Use when performing deep code analysis for correctness, safety, quality, or to simplify recently modified code. Trigger on feature completion, unpushed commits, or before merging significant changes. - Skill: `tomevault-io/review-7` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/review-7` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/review-7/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/review-7 --- # Review (Audit & Refine) Analyze code for correctness, safety, and efficiency. This skill supports two modes: **Report** (analysis only) and **Refine** (apply improvements). ## Core Mandates - **Evidence-First:** Establish a baseline by running tests BEFORE the review. - **Single-Pass:** Evaluate all categories (Correctness, Complexity, Quality, Efficiency, Cleanliness) in one pass. - **Standards:** Strictly follow the `Code Standards` table in `CLAUDE.md`. - **Mode Selection:** - **Report:** Output a categorized report (`ERROR`, `WARN`, `NIT`). End with `Verdict: LGTM`, `LGTM with nits`, or `Needs work`. - **Refine:** Apply targeted, behavior-preserving changes to improve clarity and reduce technical debt. ## Execution Workflow 1. **Scope:** Detect modified files, feature branch diffs, or staged changes. 2. **Baseline:** Run `build` and `test` to establish a "Green" starting state. 3. **Analyze:** Apply the `Code Standards` (referencing `CLAUDE.md`). 4. **Action:** - **Report Mode:** Generate findings formatted as `[SEVERITY] file:line - Issue -> Fix`. - **Refine Mode:** Edit files surgically. Use language-expert skills (`rust-expert`, etc.) for idiomatic patterns. 5. **Verify:** Run tests/linter after refinement to ensure behavior is unchanged. ## Anti-Rationalization | Excuse | Reality | | :--- | :--- | | "Tests passed, so it's fine." | Reviews catch architectural flaws and security risks that tests miss. | | "Refining might break it." | Small, behavior-preserving edits with Red-Green verification are safe and essential. | | "Subagents are always better." | For small or local changes, the main agent is more efficient. Use subagents for context isolation. | --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/nijaru) — claim your Tome and manage your conversions.