AI Security Code Review Agent

An agent that scans every PR for SQL injection, XSS, secret leaks, auth bypass, and 12+ OWASP Top-10 patterns — before merge.
STATUS · LIVE
v1

01 / Invocation - the dream begins

AI Security Code Review Agent

An agent that scans every PR for SQL injection, XSS, secret leaks, auth bypass, and 12+ OWASP Top-10 patterns — before merge.

// the dream begins · roots anchor data · intelligence awakens

AI Security Code Review Agent on Taskade Genesis. Every PR gets a security pass before human review: SQLi, XSS, hardcoded secrets, auth bypass, broken access control, insecure deserialization, and 12+ OWASP Top-10 patterns.

What's Included

  • Secret Scanner: AWS / GCP / OpenAI / Stripe key patterns + entropy check
  • OWASP Top-10 Pass: Reads diffs for the 12 most common vuln patterns
  • Auth Boundary Watch: Flags any new endpoint that lacks an auth middleware
  • Dependency Risk: Reads package.json / pyproject.toml diffs against known-vulnerable versions
  • Approve / Block: Posts a structured PR comment + sets a status check

How To Use

  1. Clone this app
  2. Connect GitHub via webhook
  3. The security agent reads every PR diff and posts findings as a PR comment
  4. Configure block thresholds (high-severity → block merge; medium → warn-only)