webcmd-adapter-author

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow can ingest OUTSIDER-authored free text via webcmd browser analyze/open/network fetching public web pages and extracting their readable HTML/text (e.g., arbitrary site content from the target URL) into the agent’s LLM context, creating an indirect prompt-injection risk.