fto-search
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests technical proposals provided by users, which serves as a surface for indirect prompt injection. While the skill includes logical checks to ensure only a single technical scheme is analyzed at once, it does not define specific boundary markers or sanitization procedures to distinguish between data and potentially malicious instructions within the proposal.
- Ingestion points:
technical_proposalparameter inSKILL.md. - Boundary markers: Not explicitly defined in the prompt templates.
- Capability inventory: The skill orchestrates multiple MCP tools (
fto-search-mcp) capable of query generation, patent database searching, and automated report creation. - Sanitization: No explicit sanitization or escaping of the user-supplied text is described before processing.
Audit Metadata