mcmc-sampling-stan
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- Unverifiable Dependencies & Remote Code Execution (SAFE): The skill suggests installing the rstan package from the official CRAN mirror (https://cloud.r-project.org/), which is a trusted source for R software.
- Command Execution (SAFE): System commands are limited to environment and dependency verification (which, dpkg) and do not perform any destructive or unauthorized actions.
- Dynamic Execution (SAFE): While the skill involves runtime compilation of Stan models, this is the core functionality of the library and is executed locally on user-provided model code.
Audit Metadata