margin-trading
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is authored by the official vendor (Binance) and targets legitimate vendor infrastructure at api.binance.com.
- [CREDENTIALS_UNSAFE]: No actual credentials are exposed. The skill uses clearly marked placeholders like 'your_mainnet_api_key' and 'your_mainnet_secret' within its examples and documentation.
- [COMMAND_EXECUTION]: The skill includes instructional shell script examples utilizing 'curl' and 'openssl' for the purpose of demonstrating how to sign API requests. These scripts are provided for educational purposes and are not executed automatically.
- [DATA_EXFILTRATION]: Network interactions are restricted to the vendor's own official domain (api.binance.com). The skill includes explicit instructions for the agent to never send credentials to third-party sites.
- [PROMPT_INJECTION]: The skill includes robust 'Agent Behavior' guidelines that instruct the agent to mask secrets, confirm mainnet transactions with the user, and maintain strict credential privacy, mitigating risks of instruction override.
Audit Metadata