🤖 PR-Agent Pro — Automated AI Pull Request Reviewer & Security Auditor

An enterprise-grade GitHub Action, Model Context Protocol (MCP) Server, and CLI tool that automatically audits pull request code diffs, detects security flaws & secret leaks, posts inline 1-click suggested fixes, and generates semantic changelogs.

⚡ Key Capabilities

• 🛡️ Zero-Day Security & Secret Leak Scanner: Scans modified lines for leaked Stripe/AWS/GitHub API keys, plaintext passwords, SQL injection vulnerabilities, and command injection risks.

• 📝 Actionable 1-Click Code Review: Posts inline GitHub review comments containing suggestion code blocks that developers can apply directly from the GitHub UI with one click.

• 🤖 Semantic PR Summary & Changelog: Automatically synthesizes human-readable pull request overviews, risk scores, and release note bullets.

• 🔌 Model Context Protocol (MCP) Server: Equips Claude Desktop, Cursor IDE, and Antigravity agents to review local git diffs before pushing!

Related MCP server: aigently

🚀 GitHub Action Quickstart (2 Minutes)

Add this workflow to your repository at .github/workflows/ai-pr-reviewer.yml:

name: AI Pull Request Reviewer

on:
  pull_request:
    types: [opened, synchronize, reopened]

jobs:
  ai-review:
    runs-on: ubuntu-latest
    permissions:
      contents: read
      pull-requests: write
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4

      - name: Run PR-Agent Pro Reviewer
        uses: meanusarcanus/ai-pr-reviewer-action@master
        with:
          github_token: ${{ secrets.GITHUB_TOKEN }}
          severity_threshold: "Medium"

🛠️ Model Context Protocol (MCP) Tools

Connect to Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "ai-pr-reviewer": {
      "command": "python3",
      "args": ["/path/to/ai_pr_reviewer_action/mcp_server.py"]
    }
  }
}

📦 Python SDK & CLI Usage

pip install ai-pr-reviewer

from ai_pr_reviewer import review_diff, audit_security, generate_changelog

diff = """
--- a/db.py
+++ b/db.py
@@ -10,3 +10,4 @@
+    query = f"SELECT * FROM users WHERE email = '{user_email}'"
+    cursor.execute(query)
"""

# Audit security
security = audit_security(diff)
print(f"Score: {security['security_score']}/100 | Grade: {security['security_grade']}")

# Run code review
review = review_code_diff(diff)
print(f"Verdict: {review['verdict']}")

📄 License

MIT License. Created by Meanus Arcanus.