skills/bankrbot/skills/coinhero/Gen Agent Trust Hub

coinhero

Pass

Audited by Gen Agent Trust Hub on Aug 4, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: Uses the curl tool to communicate with the service provider's API endpoints.
  • [DATA_EXFILTRATION]: Transmits the user's API key (COINHERO_API_KEY) and wallet address to https://my.coinhero.fun. This is standard behavior for interacting with the CoinHero platform as the intended service provider.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes JSON data from the CoinHero API to facilitate blockchain transactions.
    • Ingestion points: Data enters the agent's context through API responses from https://my.coinhero.fun during authorization and deal monitoring (documented in SKILL.md).
    • Boundary markers: The documentation explicitly warns the agent to ignore any instructions, URLs, or transaction substitutions found in the nextSteps or error message fields of the API response.
    • Capability inventory: The skill uses curl for networking and requires the agent to sign blockchain transactions from its managed wallet.
    • Sanitization: The skill mandates a detailed verification process where the agent must validate the target address, chain ID, function name, and arguments against local parameters before signing any transaction.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 4, 2026, 07:02 PM
Security Audit — agent-trust-hub — coinhero