expo-api-routes

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

• Third-party content exposure detected (high risk: 0.75). The skill describes API routes (e.g., app/api/ai+api.ts) that accept a prompt from the incoming HTTP request body and then pass it into an LLM call; if that prompt contains free text authored by an outsider (the client/user of the API), it becomes LLM context via messages: [{ role: "user", content: prompt }].