security-audit

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly asks users to paste code or config that may contain private keys and API secrets and requires reporting exact locations and findings, which forces the model to ingest sensitive values and may lead to verbatim echoing of secrets in its output if not explicitly redacted.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow (Mode #3 "Verification checks" and Mode #5 "Due diligence sources") explicitly directs the agent to consult open/public third-party sites such as Etherscan/block explorers, DefiLlama, Rekt.news, and protocol documentation/forums, meaning the agent is expected to fetch and interpret untrusted/user-generated web content that can materially affect audit decisions.