sample-correlation-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill suggests installing required R packages from the official Comprehensive R Archive Network (CRAN) at https://cloud.r-project.org, which is a well-known and trusted service.
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection via processed data files.
- Ingestion points: The read_data function in scripts/utils.R loads external CSV, TXT, or TSV files into the R environment.
- Boundary markers: None are present to distinguish data from instructions.
- Capability inventory: The skill can create directories (dir.create) and write files (write.csv, write.table, writeLines) in scripts/run_analysis.R and scripts/utils.R.
- Sanitization: The resolve_variable_pair function in scripts/functions.R explicitly converts variable data to numeric types (as.numeric) before analysis, which prevents the data from being interpreted as code or instructions by the statistical engine.
- [COMMAND_EXECUTION]: The skill executes local R scripts using the Rscript command. It employs the optparse library to define and validate specific command-line arguments, reducing the risk of arbitrary command injection.
- [DATA_EXFILTRATION]: The skill writes R session information to the output directory as session_info.txt. This is a standard data science practice for reproducibility and does not involve exfiltrating sensitive user data to external locations.
Audit Metadata