AI Pull Request Reviewer

What it does

Sets up an automated AI review on every pull request in your GitHub repository via GitHub Actions. It fails a required check when it finds a real defect, can auto-merge clean low-risk changes if you enable it, and freezes anything touching the paths you marked dangerous until a human approves.

When to use it

• When you have a team or several contributors sending pull requests regularly.
• When you want an automatic filter before a human reviews the code.
• When critical files (payments, auth) shouldn't be touched without approval.

Steps

1. Connect the reviewer to your GitHub repository using your own credential.
2. Define the paths you consider sensitive or dangerous.
3. Every pull request triggers an automatic review in GitHub Actions.
4. If it finds a real defect, the required check fails and blocks the merge.
5. Clean low-risk changes can auto-merge if you enable that option.
6. Any change touching dangerous paths stays frozen until a human approves it.

The skill

---
name: ai-pr-reviewer
description: Installs a blocking AI code reviewer that fails real defects and freezes changes to paths you mark as dangerous.
---

# AI Pull Request Reviewer

## 1. What it does

- Sets up an automated AI review on every pull request in your GitHub repository via GitHub Actions. It fails a required check when it finds a real defect, can auto-merge clean low-risk changes if you enable it, and freezes anything touching the paths you marked dangerous until a human approves.

## 2. When to use it

- When you have a team or several contributors sending pull requests regularly.
- When you want an automatic filter before a human reviews the code.
- When critical files (payments, auth) shouldn't be touched without approval.

## 3. Step by step

- Connect the reviewer to your GitHub repository using your own credential.
- Define the paths you consider sensitive or dangerous.
- Every pull request triggers an automatic review in GitHub Actions.
- If it finds a real defect, the required check fails and blocks the merge.
- Clean low-risk changes can auto-merge if you enable that option.
- Any change touching dangerous paths stays frozen until a human approves it.

## 4. Safety rules

- Safety rules: everything ships as a draft first; nothing is sent, published, or paid for without your explicit approval; no data, numbers, or results are ever invented; no credentials or secrets are typed into chat.

## 5. Best practices

- Always review the full draft before approving any send, publish, or payment.
- If real information (data, numbers, results) is missing, ask for it to be flagged as pending instead of invented.
- Keep credentials and keys out of chat; set them in environment variables or your secrets manager.

## 6. GoHighLevel integration

- Where relevant, results can be saved as contacts, tags, tasks, or notes inside your GoHighLevel sub-account.
- No automation inside GHL (workflow, campaign, trigger) activates without your review first.

For Claude Code, save it as ~/.claude/skills/ai-pr-reviewer/SKILL.md. In Cursor, ChatGPT or another assistant, paste it as project instructions.