triage-pr-feedback

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill automates PR triage but explicitly instructs excluding automated checks like Sonar and Snyk, which suspiciously weakens security review without clear malicious intent.

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

• Third-party content exposure detected (medium risk: 0.30). The workflow requires fetching and reading a user-provided GitHub pull request URL, including inline review threads and review summaries authored by humans or Copilot, which are outsider-authored free text the agent ingests at runtime.