implement

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

• Third-party content exposure detected (medium risk: 0.65). This skill can ingest outsider-authored free text when the required workflow step “Resolve and read the input” handles an explicit GitHub issue (step 3: “For an issue, the body becomes the input”) and then uses that content to derive implicit tasks and inform in-session self-review and reports, creating an indirect prompt-injection path via the issue body.