fund-risk-analyzer
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external CSV data which creates a surface for indirect prompt injection.\n
- Ingestion points: The
read_csv_navfunction inscripts/etf_screener.pyreads data from the file path provided via the--inputargument.\n - Boundary markers: There are no explicit delimiters or instructions for the agent to ignore potentially malicious content within the processed CSV files.\n
- Capability inventory: The script is limited to local file reading (
--input) and writing (--output). It does not have network access or the ability to execute system commands.\n - Sanitization: The script performs strict numeric parsing (casting to float), which prevents executable code in data cells. However, CSV headers are processed as strings and printed in the report, which could be used to expose the agent to unauthorized instructions.
Audit Metadata