Switch and manage Ollama models

Pull and manage Ollama models for minions

When setting up or switching between different Ollama code/reasoning models with different speed/quality tradeoffs (0.5b nano → 14b large).

• · Model preset name (nano/small/medium/large)
• · Optional: specific model tag

• · List of installed models
• · Model info
• · Storage usage breakdown

• · ollama list/pull/rm/show CLI

Ollama installed; network access for model pulls

• · Ollama not running → command fails
• · Disk full → model pull fails