exploiting-websocket-vulnerabilities

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes command-line utilities such as curl, wscat, and websocat to probe WebSocket endpoints and verify handshake behaviors.
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of legitimate security tools from trusted registries, including npm (wscat), Cargo (websocat), and PyPI (websockets, requests).
  • [DATA_EXFILTRATION]: The skill includes an educational proof-of-concept for Cross-Site WebSocket Hijacking (CSWSH) that demonstrates how sensitive data could be exfiltrated to an external domain (attacker.example.com) if the vulnerability is present.
  • [SAFE]: All security-sensitive actions, such as sending exploit payloads including path traversal or command injection strings, are clearly documented as part of an authorized vulnerability assessment workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 04:13 PM
Security Audit — agent-trust-hub — exploiting-websocket-vulnerabilities