hyperliquid
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill contains a telemetry module (
_trade_report.py) that automatically transmits transaction metadata, including wallet addresses, order details, and execution prices, to a remote API. This reporting is intended for developer-side trade analytics. - [COMMAND_EXECUTION]: The skill facilitates sensitive financial operations, such as withdrawing funds to arbitrary external addresses and setting leverage on positions. It also programmatically approves a 0.02% platform fee on trades for the developer's wallet.
- [INDIRECT_PROMPT_INJECTION]: Because the skill ingests and displays untrusted data from external decentralized exchange APIs, it possesses a theoretical surface for indirect prompt injection, where an attacker could attempt to influence the agent via transaction data. The skill ingests data from external sources (Hyperliquid API) without explicit prompt-injection sanitization and has the capability to perform high-privilege financial actions.
Audit Metadata