tooluniverse-cell-line-profiling

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads large scientific datasets (CRISPRGeneEffect.csv and Model.csv) from depmap.org, a well-known research repository maintained by the Broad Institute of MIT and Harvard.
  • [COMMAND_EXECUTION]: The instructions direct the agent to write and execute Python scripts via Bash to perform statistical computations, data processing, and visualization (using pandas, scipy, and matplotlib) on retrieved biological data.
  • [DYNAMIC_EXECUTION]: The workflow involves generating and running code at runtime to analyze model dependencies and drug sensitivity profiles based on user-supplied queries.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from multiple external biological databases (DepMap, Cellosaurus, PharmacoDB, COSMIC) which serves as an ingestion surface for untrusted content.
  • Ingestion points: Data retrieved through tool calls and external CSV downloads from the DepMap portal.
  • Boundary markers: The instructions include tool parameter validation tables but lack specific delimiters for handling external data in the final report context.
  • Capability inventory: The skill possesses file system write access for caching, network access for downloading datasets, and shell execution capabilities for analysis.
  • Sanitization: No explicit sanitization or filtering of the retrieved scientific metadata is specified before it is incorporated into the agent's reasoning process.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 06:31 AM
Security Audit — agent-trust-hub — tooluniverse-cell-line-profiling