skills/datadrivenconstruction/ddc_skills_for_ai_agents_in_construction/csv-handler/Gen Agent Trust Hub
csv-handler
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows secure coding practices for its intended purpose of construction data processing. All identified behaviors are consistent with the described functionality.
- [COMMAND_EXECUTION]: The skill performs standard filesystem operations (reading/writing CSV files) using the pandas library and Python's pathlib. These operations are scoped to the data processing tasks requested by the user and do not include arbitrary command execution.
- [DATA_EXPOSURE]: No hardcoded credentials or sensitive system file paths were found. The skill only interacts with data provided by the user or generated during the CSV processing workflow.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface through CSV file processing (SKILL.md, ConstructionCSVHandler.read_csv). While it lacks explicit boundary markers for the data content, the processing logic is strictly programmatic (data cleaning and statistical summary) and does not interpret CSV content as instructions for the agent.
Audit Metadata