bingx-swap-trade

Pass

Audited by Gen Agent Trust Hub on Oct 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input (symbols, prices, quantities) to construct and execute financial transactions. While it implements comprehensive validation logic, this creates a surface for indirect prompt injection.\n
  • Ingestion points: User-supplied parameters for orders, leverage, and margin settings (SKILL.md).\n
  • Boundary markers: Requires explicit user confirmation ("CONFIRM") for production environment and uses strict regex/enum validation.\n
  • Capability inventory: Authenticated trading operations including placing/canceling orders and managing positions (SKILL.md, api-reference.md).\n
  • Sanitization: Implements strict character filtering (rejecting &, =, ?, #, \\n) and enforces typed serialization of parameters.\n- [COMMAND_EXECUTION]: The skill provides template code (fetchSigned) for the agent to execute network requests to external API endpoints. This code handles HMAC signing and domain failover.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 10, 2026, 10:52 AM