ai-tells-review

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

• Third-party content exposure detected (high risk: 0.70). This skill’s required workflow is a “read the artifact end to end” and then generate a rewrite report; that means it ingests the user-provided artifact prose into the LLM context, which could be outsider-authored free text (e.g., public web content or other people’s docs) depending on what the operating user supplies.