tooluniverse-precision-medicine-stratification
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted patient profiles, including genomic and clinical data, which introduces an attack surface for indirect prompt injection.
- Ingestion points: Patient genomic profiles, clinical biomarkers, and therapeutic data entered via prompts (SKILL.md).
- Boundary markers: The skill instructions do not specify the use of delimiters or provide instructions to disregard embedded instructions within the ingested patient data.
- Capability inventory: The skill is capable of generating and executing Python scripts using Bash and possesses extensive network capabilities via the ToolUniverse toolset (SKILL.md, TOOLS_REFERENCE.md).
- Sanitization: There are no explicit instructions for sanitizing or validating the ingested patient information before it is processed by the agent.
- [DYNAMIC_EXECUTION]: The skill uses runtime script generation and execution for its core analytical tasks.
- Evidence: The SKILL.md file contains a dedicated section "COMPUTE, DON'T DESCRIBE" which instructs the agent to "write and run Python code via Bash" to perform statistics, data processing, and analysis tasks.
Audit Metadata