options-market-making
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data from a CSV file.
- Ingestion points: The script
scripts/options_market_making_diagnostics.pyreads data from a user-providedinput.csv. - Boundary markers: The instructions for the agent do not incorporate specific boundary markers or instructions to ignore embedded commands when interpreting the
diagnostics.jsonoutput. - Capability inventory: The skill utilizes a Python script to perform statistical analysis and output results to the file system. The agent then consumes this data for reporting.
- Sanitization: The script implements partial sanitization by coercing data values to numeric and datetime types. However, it preserves the original column headers from the input CSV in the final JSON output, which allows for a potential injection surface if headers are maliciously crafted.
Audit Metadata