maker
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements secure credential handling by instructing users to utilize environment variables (e.g., PRIVATE_KEY and RPC_URL) rather than hardcoding sensitive data into scripts.
- [SAFE]: All smart contract addresses, ABI definitions, and external documentation links refer to verified and official MakerDAO and Sky Protocol resources, which are well-known and trusted entities in the decentralized finance space.
- [PROMPT_INJECTION]: The skill includes technical guidance for the agent to prevent logic errors related to the protocol's unique unit systems (wad, ray, rad). While these instructions are presented as 'non-negotiable' to ensure technical accuracy, they are benign and do not attempt to bypass agent safety filters or constraints.
- [SAFE]: The logic provided for processing blockchain data (e.g., vault collateralization ratios) uses structured ABIs and specific data types, effectively isolating the agent from potential indirect prompt injection risks associated with ingesting raw data from the blockchain.
Audit Metadata