alterlab-qualitative-methods
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external qualitative data such as transcripts and documents through Python scripts. This represents a standard attack surface for qualitative research tools where external content could potentially include instructions designed to influence agent behavior.
- Ingestion points: Python scripts in 'references/python_workflows.md' read text files from a data directory.
- Boundary markers: The provided script templates do not include explicit delimiters or 'ignore' instructions for the data being read.
- Capability inventory: The skill utilizes 'Bash(python:*)' for data processing and saving visualizations to the filesystem.
- Sanitization: The scripts perform basic string operations (lowercase, strip) as part of the thematic analysis process, which is standard for the use case.
- [DYNAMIC_EXECUTION]: The skill provides Python code snippets in 'references/python_workflows.md' for keyword coding, word frequency analysis, and thematic network visualization. These are executed by the agent via 'uv run python' as specified in the compatibility metadata to support research analysis workflows.
Audit Metadata