The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional markdown and static JSON templates (assets/). There is no executable logic, shell commands, or script files provided.
[PROMPT_INJECTION]: The instructions focus purely on system design and implementation workflows for real-time communication. No attempts to override safety filters, bypass instructions, or extract system prompts were detected.
[DATA_EXFILTRATION]: There are no network-related commands (like curl or wget) or references to sensitive file paths (like credentials or SSH keys). The allowed-tools field in the frontmatter is empty, restricting the agent's ability to use external tools.
[COMMAND_EXECUTION]: No shell commands or dynamic execution patterns (eval, exec, or !command syntax) are present in the provided files.
[REMOTE_CODE_EXECUTION]: There are no dependencies or external scripts fetched at runtime.

websocket-engineer