Audit AI applications and agents for security risks before commit, PR, merge, or deployment. Deterministic findings with coverage and evidence receipts.

What changed in your stack since the model's cutoff, error fixes, and security-scanned skills/MCPs