All 94 skillsWorkflow12 of 15

stryx:local-ai

When running AI models on your own graphics card: Ollama, LM Studio, ComfyUI.

/stryx:local-ai

Inside its piece: 15 specks, one for each file on its shelf. Its fourteen group-mates glow on the open book behind it. The light splits toward indigo, the band of Workflow.

In numbers

counted 2026-10-12 in the installed set
  • 6reference files
  • 7helper scripts
  • 15files in the skill
  • 157KB on disk
  • 7chapters

Use it when

in the skill’s own words

When running, checking or using AI models on the user's own GPU

    • Ollama
    • LM Studio (lms)
    • a ComfyUI server
    • GGUF or MoE models
    • llama.cpp or vLLM
  • What is loaded, how much VRAM is free or what nvidia-smi says
  • Loading, unloading or freeing GPU memory
  • Handing bulk reading, summarizing, classifying, extracting or embedding to local models to save Claude tokens ("use the GPU", "use local models", "use ollama", "use the vram")
  • Choosing a local model or whether its license allows business use
  • rembg background removal
    • installing
    • updating or troubleshooting NVIDIA NemoClaw
    • OpenClaw or OpenShell sandboxed agents on Linux
    • WSL2
    • DGX Spark or DGX Station (port 8080/18789 conflicts, dashboard tokens, tunnel access, sandbox skills, rebuilds)

Not for

another skill does it

On its shelf

15 files, 157 KB
  • scripts/7
  • references/6
  • SKILL.md1
  • evals/1

Inside the book

7 chapters, by title only
  1. The laws of a job a local model writes
  2. The stack (labeled example: the owner's home PC)
  3. The procedure
  4. Rules
  5. What a pass looks like
  6. NemoClaw in one paragraph
  7. Scripts

Works with

the skills it names most

Claude loads it by itself when the moment comes; by hand, type the line at the top. It comes with the whole set: Installs all 94 skills, 32 agents and the brain program as one set. The skill’s own text is private source and is not shown here.