Release risk, dependency drift, test posture, rollback readiness
Security release gates
Create a durable passport for AI-assisted ship/no-ship recommendations so reviewers can prove which release decision was approved.
Use cases
Decision provenance for security and AI governance teams
Decision Passport is designed for workflows where a recommendation is useful only if stakeholders can later prove what was reviewed, approved, revoked, or superseded.
Severity, exploit evidence, internet exposure, patch availability
Vulnerability triage
Turn urgent triage into a verifiable decision record without publishing private exploit details or customer impact notes.
Model version, policy version, reviewer action, decision status
AI governance
Give risk, legal, and audit teams a reliable way to inspect whether an AI-assisted decision changed after review.
Tool risk, policy match, sensitive action, human approval requirement
Agentic policy control
Use decision passports for agentic workflows that need policy gates before sensitive tool use or external action.
Common operating pattern
The agent prepares a structured recommendation, the backend hashes the canonical object, and a reviewer makes the accountable lifecycle decision. The public verifier checks integrity without receiving private evidence.
Not a data dump
The product is intentionally narrow. It does not publish prompts, raw logs, customer identifiers, source code, or private vulnerability details. It publishes proof that a governed decision object still matches the approved passport.