container-security-skill

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes explicit examples that place credentials directly on the command line (e.g., --username/--password ) and uses placeholders for tokens, which would require the LLM to include secret values verbatim if filled in and therefore poses a high exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflow explicitly imports and scans public images and binaries from third-party sources (e.g., az acr import --source docker.io/:, curl from GitHub releases, and Trivy scans of registry images), which are untrusted/user-generated content that the agent is expected to read and use to make remediation and build decisions.