pmfi-parbitrage

Pass

Audited by Gen Agent Trust Hub on Jul 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill hardcodes critical blockchain infrastructure, including the PMFI Vault address (0xd1ccbc2aa6e2f41817b62448089d4125e62df4fb), the Base USDC address, and the official platform API endpoint.
  • [SAFE]: A transaction allowlist implemented in the execution script restricts agent interactions to specific function selectors (requestDeposit, requestRedeem, and approve), preventing unauthorized contract interactions.
  • [SAFE]: Extensive pre-execution validation is performed, including network ID verification, contract bytecode checks, vault state (paused/shutdown) monitoring, and transaction simulation before submission.
  • [SAFE]: Security guardrails explicitly instruct the agent to treat external content such as RPC responses, tool outputs, and documentation as untrusted data that cannot override hardcoded configurations.
  • [SAFE]: The skill enforces a mandatory risk disclosure and direct user confirmation flow for all financial actions, ensuring transparency and user control.
  • [SAFE]: Local credential management is handled through standard platform configuration files and environment variables, with network usage restricted to the reviewed official API domain.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 5, 2026, 02:11 PM
Security Audit — agent-trust-hub — pmfi-parbitrage