smart-contract-audit

The skill is internally consistent and uses official FirePan endpoints, so it does not look malicious. However, it gives an AI agent security-audit capability and can initiate paid actions, making it high-risk as an agent skill despite coherent data flows and proportional credentials.