biopharma-fto-analysis
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process significant amounts of external data, including patent text, drug intelligence, and user-provided technical specifications, which constitutes an indirect prompt injection attack surface.
- Ingestion points: User-defined technical parameters (product name, domain, description) and external data retrieved through integrated search tools (e.g., patent full-texts, drug databases).
- Boundary markers: While explicit delimiters are not defined, the instructions mandate clear structural separation between facts, machine-generated content, and human interpretation.
- Capability inventory: The workflow utilizes various specialized retrieval tools (e.g.,
ls_sequence_search_submit,ls_structure_search,mcp_patent-search__patsnap_search) to fetch and analyze complex data. - Sanitization: The methodology requires manual verification of legal statuses against official registers and professional review of all significant risk findings, which mitigates the risk of automated obedience to embedded instructions.
Audit Metadata