skills/npc-live/clawfirm/binance/Gen Agent Trust Hub

binance

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @binance/binance-cli package from the npm registry during the setup process to enable its core functionality.
  • [COMMAND_EXECUTION]: The skill executes financial operations through the binance-cli tool. These include high-impact actions like placing orders and managing withdrawals.
  • Evidence: Reference files like references/wallet.md and references/spot.md document commands for withdraw and new-order respectively.
  • [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it processes data retrieved from the Binance API. This risk is managed through instructional guardrails requiring human oversight.
  • Ingestion points: Market data and account status endpoints documented in reference files like references/alpha.md and references/spot.md.
  • Boundary markers: The skill includes a 'Notes' section in SKILL.md advising users to confirm production transactions and provides specific security rules in references/auth.md.
  • Capability inventory: The agent has access to financial operations (trading, withdrawals) and the ability to send signed requests to arbitrary URLs via binance-cli request.
  • Sanitization: The skill relies on manual user verification (typing 'CONFIRM') before executing sensitive production actions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 06:07 AM
Security Audit — agent-trust-hub — binance