deep-research

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

• Third-party content exposure detected (high risk: 0.85). The required workflow explicitly uses runtime web search and scraping via outsider-authored public web content (e.g., firecrawl_search/firecrawl_scrape and crawling_exa fetch and ingest arbitrary third-party pages’ body text into the agent for synthesis), which is classic indirect prompt-injection exposure.