bingx-coinm-ws-account
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No malicious prompt injection patterns were found. The skill includes explicit safety instructions that strengthen the agent's behavior, such as 'NEVER return code to the user' and 'GET-only execution' for the agent's own operations.
- [DATA_EXFILTRATION]: The skill interacts exclusively with official BingX API and WebSocket endpoints (bingx.com). No unauthorized external network calls or sensitive data exposure patterns were identified.
- [COMMAND_EXECUTION]: The provided TypeScript code is for processing WebSocket streams and GZIP decompression using standard libraries. It does not contain any arbitrary command execution or shell injection vulnerabilities.
- [EXTERNAL_DOWNLOADS]: The skill uses the 'pako' library for decompression, which is a standard industry dependency for GZIP handling. No untrusted remote scripts or installers are referenced.
Audit Metadata