reviewing-api-reference

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime path ingests the “finished, published API reference” and the “handed-in upstream api-spec” as readable text for Step 1/2 checks; the reference content is outsider-authored (a producer other than the operating user), so it can contain free text that the LLM processes (indirect prompt injection risk via the reviewed document).