r-bayes
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill contains templates that ingest untrusted data frames into the agent's context for statistical processing.
- Ingestion points: R code snippets in SKILL.md use variables analysis_data, model_data, and data as inputs for modeling functions.
- Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded instructions for the processed data.
- Capability inventory: The skill uses the brm function which invokes a C++ compiler via cmdstanr, and file system writes for model caching.
- Sanitization: There is no evidence of data sanitization or validation before the data is passed to the Bayesian modeling engine.
- [DYNAMIC_EXECUTION]: The skill uses the brms library, which generates and compiles Stan code into binary executables at runtime.
- Evidence: The brm() call with backend = 'cmdstanr' triggers runtime C++ compilation of statistical models.
- Context: While this is a primary feature of the tool, it represents a dynamic execution vector where generated source code is compiled and run on the host system.
Audit Metadata