desk-incident-response

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data from the Hyperliquid exchange (such as fill records and order statuses) to inform agent-driven incident responses. Acting on instructions or data derived from third-party APIs constitutes a surface for indirect prompt injection.\n- Ingestion points: Exchange API responses including orderStatus, openOrders, userFills, and clearinghouseState referenced across playbooks.\n- Boundary markers: There are no documented boundary markers or instructions to isolate external data from the agent's decision-making logic.\n- Capability inventory: The skill allows the agent to propose order cancellations, risk reduction tickets, and new position entries, typically subject to human approval.\n- Sanitization: No sanitization or validation of the content of the external exchange data is specified in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 09:33 AM
Security Audit — agent-trust-hub — desk-incident-response