strava

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to call Strava API endpoints (e.g., /activities//comments, /activities, /clubs//activities) which return user-generated, public content that the skill is expected to fetch and parse as part of its workflow, allowing that untrusted content to influence subsequent actions (e.g., updates/uploads).