browser-to-api

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text is ingested from the offline browser-trace capture files (e.g., cdp/network/requests.jsonl, cdp/network/responses.jsonl, and optional browse network request.json/response.json bodies) via scripts/load.mjs → readJsonl()/loadBrowseNetworkBodies(), and then propagated through the pipeline into emitted artifacts (and thus into any LLM context that later reads those artifacts).