generate-codebook
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script (
scripts/generate_codebook.py) to analyze datasets. The script operates locally using the pandas library and does not require elevated privileges. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided datasets (CSV, Excel, etc.), which represents a potential injection surface.
- Ingestion points: The
read_tablefunction inscripts/generate_codebook.pyreads various tabular formats. - Boundary markers: The output codebook is structured (JSON/Markdown) but does not include explicit delimiters for data values.
- Capability inventory: The skill can read and write files and execute its own scripts; it lacks network access.
- Sanitization: Data is parsed using standard library functions without additional sanitization filters.
Audit Metadata