desk-post-trade-review
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external exchange records, such as order fills, historical status, and funding payments. While these sources could theoretically be manipulated to include adversarial instructions, the skill uses the data for specific mathematical computations (slippage, fees, PnL) and structured reporting, which aligns with standard agent operational procedures.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads from and writes to the
/workspace/trading-desk/directory to maintain journals and incident reviews. It processes trade identifiers and pricing data, but no evidence was found of hardcoded secrets or instructions to transmit this data to non-whitelisted external destinations. Communication of review results is directed toward internal roles (Desk Lead) and the user.
Audit Metadata