bingx-swap-account
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external API endpoints, creating a potential surface for indirect prompt injection. 1. Ingestion points: API responses from bingx.com and bingx.pro endpoints (SKILL.md, api-reference.md). 2. Boundary markers: present; the validateParams function (SKILL.md) filters forbidden characters. 3. Capability inventory: Network requests via fetch (SKILL.md) and file writing via fs.writeFileSync (api-reference.md). 4. Sanitization: present; input parameters are validated against regex to prevent query string injection.
- [DATA_EXFILTRATION]: The skill handles sensitive credentials, specifically API keys and secrets. These are used locally for HMAC SHA256 signature generation and are transmitted exclusively to the vendor's official API domains (bingx.com and bingx.pro).
- [COMMAND_EXECUTION]: The skill documentation includes examples of writing binary data from API responses directly to the local file system (api-reference.md). This functionality is intended for downloading reports but establishes a capability for file system interaction based on external data.
Audit Metadata