Hostile code reviewer
Point it at recent changes. Get a tough review from three hostile reviewers, with each finding ranked.
Hostile code reviewer is a free AI skill for the job "Test and fix code". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by Alireza Rezvani · Thinking: high
Made by Alireza Rezvani. See the original on GitHub; we copied it as it is, with its licence.
What it does
Code review from people who want to find fault, not approve. Point it at staged changes, recent commits or one file, and get a ranked list of real problems with a clear merge verdict. An AI that read the code shares its blind spots and tends to say it looks fine. This skill breaks that habit with three hostile reviewers. The Saboteur looks for what will break in production. The New Hire asks what will confuse the next person. The Security Auditor hunts for common attack holes. Each must find something. A problem spotted by two reviewers moves up a severity level. The report ends in BLOCK, CONCERNS or CLEAN.
You give it
- The code changes, a commit range or a file path
- A note on what the change is meant to do (optional)
- Areas you worry about most (optional)
You get
- Findings from three reviewer viewpoints
- Severity rankings with duplicates merged
- A BLOCK, CONCERNS or CLEAN verdict
- Suggested fixes for each finding
Use it when
- You wrote a change yourself and need a review that will not just agree with you.
- You are about to merge a pull request and want a final tough pass.
- You want to know whether new code is easy for a teammate to maintain.
- You need a quick security check on a file before it ships.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can read your project files.