hydrex
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from official Hydrex APIs to determine transaction parameters such as reward amounts and pool efficiencies. Ingestion points include references/rewards.md, references/single-sided-liquidity.md, and references/voting.md, which fetch data from api.hydrex.fi and incentives-api.hydrex.fi. While explicit prompt boundary markers and sanitization steps for the API content are not detailed in the instructions, the data is primarily used to populate blockchain transaction parameters that are ultimately managed through the bankr tool's execution flow.
- [SAFE]: The skill follows security best practices for DeFi interactions, using official protocol contracts on the Base mainnet and referencing verified documentation and APIs. All external network requests and transaction calls are consistent with the skill's stated purpose of providing a governance and liquidity interface.
Audit Metadata