runtime-admissibility-review

Determines whether a specific AI-agent action, output, recommendation, or proposed commitment remains admissible for execution or institutional reliance under current authority, delegated scope, evidence, facts, policy, risk, escalation, and revocation conditions. Use this Skill before an enterprise or regulated AI agent executes, updates records, triggers …

Independent GitHub snapshot · Automated classification · Original documentation governs