to-tickets

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md:11 instructs the agent to fetch and read the full body and comments when the user provides a reference argument (e.g., an issue number or URL), which can include outsider-authored free text from public web pages or third-party issue/comment threads that then becomes LLM-readable context.