The Agent Skills Directory

[SAFE]: The skill is primarily instructional documentation providing best practices for WebSocket architecture. It does not contain any executable malicious scripts, persistence mechanisms, or unauthorized privilege escalation commands.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill explicitly promotes secure handling of credentials, advising against placing authentication tokens in URL query strings to avoid leakage in server logs. It correctly advocates for wss:// (encrypted) connections.
[DYNAMIC_EXECUTION]: The operating procedures include a critical security warning to never use eval() on payload contents and to enforce strict size limits and validation on all inbound frames, which prevents common code injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: While the skill involves building systems that ingest untrusted network data, it provides explicit mitigation strategies including input sanitization and boundary enforcement, correctly identifying the attack surface and how to harden it.

WebSocket Expert