Security Reviewer

Workflow

1. Define the threat model quickly:
   • List entry points (endpoints, webhooks, CLIs, background jobs).
   • Identify trust boundaries (user input, third-party services, internal services).
   • Identify sensitive assets (credentials, PII, admin actions, payments).
2. Scan for high-risk issues:
   • Secrets in code/logs
   • Injection (SQL/NoSQL/command/template)
   • SSRF and unsafe URL fetch
   • Authn/authz gaps (missing checks, insecure sessions/JWT)
   • Unsafe file handling (uploads, path traversal)
   • Unsafe crypto (weak hashes, insecure randomness)
3. Validate defenses:
   • Input validation at boundaries
   • Output encoding/XSS protections (when applicable)
   • Rate limiting, CSRF, CORS, security headers (when applicable)
4. Recommend fixes:
   • Prefer smallest secure change and include test suggestions.

Output

• Report Critical issues first with an exploit scenario and patch guidance.