offensive-osint

SUSPICIOUS/HIGH-RISK skill, not confirmed malware. Its offensive recon purpose aligns with its capabilities, but the skill materially expands an AI agent’s attack surface: active probing guidance, credential validation/forwarding, disabled TLS checks in examples, and broad installation of mutable third-party tooling. The biggest concerns are agent-enabled offensive security use, high supply-chain exposure, and handling of discovered credentials, not hidden exfiltration or deceptive publisher behavior.