disease-investigation-zhcn
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to configure a connection to official vendor-hosted endpoints at patsnap.com for pharmaceutical intelligence data transport.
- [COMMAND_EXECUTION]: Provides instructions for the user to execute setup commands to install necessary MCP services, which is a standard integration step for the agent platform.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it synthesizes reports from untrusted external sources.
- Ingestion points: Processes external academic literature, epidemiological data, and web search results via ls_X_search and network search tools.
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when interpolating fetched data.
- Capability inventory: The agent uses search and retrieval tools to collect data across multiple domains (patents, trials, web) and synthesizes findings into a final report.
- Sanitization: There is no explicit sanitization of external content before it is included in the agent's context.
Audit Metadata