API-BACKED FULL LOOP
SELF-AUDIT ONLYSOURCE-ONLY · SELF-AUDIT
Find the bug.Patch it. Re-analyze it.
Add source files from an application you own, confirm authorization, and the API-backed Deadbolt engine can map the trust boundaries, run four focused hunts, prepare reviewable patches, and rerun each affected hunt against the patched clone. It requires a configured provider; the keyless $deadbolt Codex Skill is a separate read-only audit and does not run this loop.
CHOOSE WHAT TO SCAN
Checking whether the API-backed engine is available on this deployment.
API-BACKED HUNT → PATCH → RE-ANALYZE PIPELINE
No analysis run yet
Intake
—Threat model
—Four hunts
—Human stakes
—Focused patches
—Hunt re-analysis
—Add source files and confirm ownership to start the threat model, four focused hunts, isolated patches, and affected-hunt re-analysis.
DEFENSIVE BOUNDARY
Source code only. No third-party probing. No exploit payloads.