baseline-security-audit

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill instructs the agent to fetch and follow external instructions from a remote URL, introducing a dynamic instruction-loading mechanism without malicious intent.

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

• Third-party content exposure detected (low risk: 0.10). The workflow fetches dynamic instructions from a public raw GitHub URL controlled by outsiders.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill fetches behavior-governing instructions from an external GitHub repository with weak provenance, creating a high-tier operational dependency.