The Agent Skills Directory

[SAFE]: The skill functions as a technical guide for LLM deployment and serving. All provided examples follow standard developer practices for the mentioned tools.
[COMMAND_EXECUTION]: The skill includes shell commands for starting model servers and compiling local binaries (e.g., llama.cpp). These commands are intended for local environment setup and do not involve suspicious network exfiltration or privilege escalation.
[EXTERNAL_DOWNLOADS]: Includes commands that trigger model downloads from official repositories (e.g., ollama pull and vLLM model loading from Hugging Face). These are expected behaviors for LLM deployment skills and target well-known, trusted services.

llm-deployment