hoc-pattern

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly requires the withLoader HOC to fetch data from arbitrary public API endpoints (e.g., the example URL "https://dog.ceo/api/..." and instructions saying the HOC "fetches the data from the API endpoint that we pass"), so the agent will ingest and act on untrusted third‑party content at runtime.