Local AI workstation setup

[funnicus]

Ollama
Open WebUI
model pull list
systemd service config
GPU monitoring tools
context-size settings
model storage location
Python/uv environment
agent sandbox directories

dotsetup ai install
dotsetup ai pull qwen-coder
dotsetup ai check
dotsetup ai status

✓ ollama running
✓ GPU visible
✓ model directory exists
✓ enough disk space
✓ selected models installed
✓ Open WebUI container healthy

[ai.models]
coding = ["qwen2.5-coder:14b", "deepseek-coder-v2:16b"]
writing = ["gemma3:12b"]
agent = ["qwen3-coder:30b"]