desk-execution-protocol
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill establishes high-rigor operational protocols for financial transactions, including mandatory slippage checks and network validation.
- [CREDENTIALS_UNSAFE]: The instructions explicitly direct the agent to access keys from the environment and never print them to logs or chat, following security best practices for secret management.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from proposal files (
/workspace/trading-desk/proposals/) and user chat approvals. The protocol mitigates this surface through a mandatory 'Ticket integrity' checklist that requires the agent to verify that the proposal content matches the Risk Manager's approval before execution. - [COMMAND_EXECUTION]: While the skill involves executing trading commands via the Hyperliquid API, these are scoped to the intended function of the desk and are gated by multiple layers of manual and automated verification.
Audit Metadata