ai-identity-security

The provided content is not a software module; it is an adversarial prompt/payload playbook targeting AI agent security. It explicitly requests credential/token/system prompt disclosure, unauthorized actions via parameter tampering, attacker-controlled remote script execution, and multi-agent identity/trust manipulation. Because no actual npm package/library code is included, confidence in a supply-chain malware assessment is limited; the fragment is nevertheless strongly malicious in intent/content.