backlinks-domain-pages-summary

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

• Third-party content exposure detected (high risk: 0.75). The required workflow calls an external Frevana endpoint at runtime and then reads the returned HTTP response body as text (curl -o "$RESPONSE_FILE" → python3 ... read_text(...) → cat "$RESULT_FILE"), so any free-form content in that outsider-authored response can be injected into the agent’s LLM context.