tooluniverse-multiomic-disease-characterization

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions explicitly direct the agent to write and execute Python code using Bash to perform statistical analysis and data processing tasks (e.g., using pandas, scipy, statsmodels).
  • [EXTERNAL_DOWNLOADS]: The skill retrieves biological and clinical data from numerous external bioinformatics APIs and well-known services, including OpenTargets, STRING, Reactome, and the GWAS Catalog, to support its multi-layered disease characterization.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted data from multiple external API sources.
  • Ingestion points: Data enters the system from external APIs such as OpenTargets (disease descriptions, associated targets), GWAS Catalog (associations), and Enrichr (pathway data), as documented in tool-reference.md.
  • Boundary markers: The instructions lack explicit boundary markers or directives to ignore instructions that may be embedded within the retrieved scientific data.
  • Capability inventory: The skill has the capability to execute shell commands and Python scripts via Bash as specified in the 'COMPUTE, DON'T DESCRIBE' section of SKILL.md.
  • Sanitization: There are no documented steps for sanitizing or validating the content retrieved from external APIs before it is used in report generation or computational analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 11:45 PM