Skip to content
Get packs

Adversarial Code Reviewer

Runs a hostile, three-persona code review on your changes to catch blind spots a self-review misses.

Source Download skill
Repository
alirezarezvani/claude-skills
Stars
29,149
Licence
MIT
Files
1 file

What it does

When installed, the agent reviews recent code changes (uncommitted edits, the last few commits, or a specific file) through three deliberately critical viewpoints: a Saboteur looking for production-breaking bugs, a New Hire judging whether the code is understandable, and a Security Auditor checking for common vulnerabilities. Each viewpoint must report at least one issue, and the result is a severity-ranked report ending in a BLOCK, CONCERNS, or CLEAN verdict. It is meant to counter the tendency of AI assistants to approve their own code too easily.

Great for

  • Developers who want a harsher second look at code before merging a pull request
  • Teams using Claude Code who suspect AI reviews of its own output are too lenient
  • Anyone preparing security-sensitive changes such as authentication or input handling

Skip it if

  • Quick style checks or formatting questions, since the forced findings can overstate minor issues (the mandatory-findings rule may produce noise on clean code)
  • Replacing a human security audit or formal penetration test, because it is prompt-based and can miss vulnerabilities
  • Large diffs where a full read of every changed file may exceed the agent's context window

Try asking

  • “/adversarial-review to critique my uncommitted changes before I commit”
  • “/adversarial-review --diff main...HEAD before I merge this PR”
  • “/adversarial-review --file src/auth.ts and look for security holes”
  • “Give me a hostile review of the last three commits focusing on production failure modes”
Step-by-step guide: how to use Adversarial Code ReviewerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    adversarial-reviewer.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip adversarial-reviewer.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip adversarial-reviewer.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run it on a focused diff rather than a whole branch so each persona can read the full context of the changed files.
  • Treat the mandatory findings as prompts to check, not proven bugs; verify each issue before acting on it.
  • Pair it with your project's tests and linters, since the review itself does not execute code.

What's inside

  • SKILL.md

Redistributed unmodified from alirezarezvani/claude-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.