Skip to content
Back to skills

Review

ASecurity

Run code review and (optionally) security review on the current diff. Phase 7 of the eight-phase workflow. Delegates to reviewer + security agents. Multi-pass — each finding is re-verified to filter false positives (Anthropic pattern).

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 9, 2026
ai-agentsgobashsqlgitdatabasesecurity

Security analysis

A100/100

Scanned October 9, 2026

npx -y skills add qyndex/claude-init --skill review --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Review?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Review
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/qyndex-review/badge)](https://www.skillsdirectory.com/skills/qyndex-review)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: review
description: Run code review and (optionally) security review on the current diff. Phase 7 of the eight-phase workflow. Delegates to reviewer + security agents. Multi-pass — each finding is re-verified to filter false positives (Anthropic pattern).
when_to_use: Implementation + verification complete; about to /ship. User says "review", "code review", "security review", "audit", "lgtm?".
argument-hint: "[--security] [--diff <ref>]"
model: opus
allowed-tools: Read, Glob, Grep, Bash, TodoWrite
---

# Review

Two-pass review on the current diff: code-quality first, security second.

## Process

1. **Determine the diff** — default to `git diff main...HEAD`. Override with `--diff <ref>`.
2. **Delegate to `reviewer` agent** for code-quality review.
3. **Delegate to `security` agent** for security review (always if `--security`, automatic if the diff touches auth/data/network/secrets/deps).
4. **Collate findings** into a single severity-sorted list.
5. **Re-verify each finding** before reporting (anti-false-positive pass).
6. **Post results** as a structured comment (PR or chat).
7. **Block /ship** if any blocker or high remains.

## Trigger conditions for automatic security pass

- Diff touches `**/auth/**`, `**/login/**`, `**/session/**`, `**/token/**`
- Diff touches database queries (SQL files, ORM definitions, migrations)
- Diff touches network code (fetch, axios, requests, http handlers)
- Diff modifies `package.json`, `requirements.txt`, `Cargo.toml`, `go.mod`
- Diff includes new env var reads
- Diff includes user-input handlers (forms, parsing, deserialization)

## Output

```
# Review — <feature> — <date>

## Code review
### Blockers (0)
### Highs (1)
- [H-1] login.ts:42 — race condition between session lookup and update; use SELECT FOR UPDATE.

### Mediums (2) / Lows (3) / Nits (5)
- ...

## Security review
### Critical (0)
### High (0)
### Medium (1)
- [M-1] User-controlled `redirect_uri` not allowlisted (open redirect). Fix: validate against config.

## Verdict
PASS-WITH-CHANGES — address H-1 and M-1 before /ship.
```

After save: "Review complete — <blocker> blockers, <high> highs. /ship blocked." OR "Review clean. Ready for /ship."

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.