A free agent skill that checks a proposed agent action against an operator-written mandate and returns permit, deny or escalate with the governing clause, then records the decision in a hash-chained local log. Runs locally, no network, no credentials. Advisory: it checks and records, it cannot enforce.
authorization audit-log mandate ai-agents hash-chain guardrails tamper-detection llm-agents policy-check claude-code agent-safety agent-governance agent-skill agent-authorization
-
Updated
Sep 7, 2026 - Python