bio-de-visualization
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill instructs the user to install R packages from CRAN and Bioconductor. According to [TRUST-SCOPE-RULE], these are recognized as trusted external sources, making the download concern low-risk.
- [COMMAND_EXECUTION] (SAFE): Analysis of the R code snippets shows no usage of dangerous functions like
system(),exec(), orshell(). All code is focused on data processing and plotting. - [DATA_EXFILTRATION] (SAFE): There are no network requests to external domains (other than standard package managers) and no attempts to access sensitive system files like SSH keys or environment variables.
- [INDIRECT_PROMPT_INJECTION] (INFO): The skill has an ingestion surface for untrusted data (differential expression result objects). However, the capability inventory is limited to visualization (plotting), which poses negligible risk to the agent context.
Audit Metadata