canny

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md shows the agent connects to Canny via Membrane and uses actions like "list-posts", "list-comments", and "retrieve-post" to fetch user-generated feedback and comments from a third-party service (Canny), which the agent is expected to read and could influence subsequent actions.