biopharma-fto-analysis

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

• Third-party content exposure detected (medium risk: 0.30). In the required workflow, the agent ingests outsider-authored free text by fetching GitHub-untrusted (in general) patent text at runtime via mcp_patent-search__patsnap_fetch/mcp_patent-search__patsnap_search, i.e., it reads full-text patent claims and descriptions returned by external patent-search tools rather than only trusted first-party documentation.