Model Router Status

Show current configuration and available models.

Trigger

User runs /model-router:status

Instructions

1. Load Configuration

Read ~/.model-router/config.yaml

If doesn't exist:

No configuration found. Run /model-router:setup to get started.

2. Check Live Status

Run environment detection to get current state:

python3 "$(dirname "$0")/../../scripts/detect_environment.py" --json

3. Display Status

┌─ Model Router Status ────────────────────────────────────┐
│                                                          │
│ Profile: [balanced/quality/budget/local_only]            │
│ Routing: [preserve_claude/quality_first/budget_first]    │
│ Tracking: [enabled (basic/detailed) / disabled]          │
│                                                          │
├─ Subscriptions ──────────────────────────────────────────┤
│                                                          │
│ Claude Pro     [✓ Active / ✗ Not configured]            │
│ ChatGPT Plus   [✓ Active / ✗ Not configured]            │
│ Google AI Pro  [✓ Active / ✗ Not configured]            │
│                                                          │
├─ CLI Tools ──────────────────────────────────────────────┤
│                                                          │
│ Codex CLI      [✓ Ready / ⚠ Installed, not auth / ✗ Missing]
│ Gemini CLI     [✓ Ready / ⚠ Installed, not auth / ✗ Missing]
│                                                          │
├─ Local Models (Ollama) ──────────────────────────────────┤
│                                                          │
│ Status: [✓ Running / ✗ Not running / ✗ Not configured]  │
│ Host: [localhost:11434 / custom host]                    │
│ Models:                                                  │
│   • llama3.3:70b      [✓ Available]                     │
│   • deepseek-r1:70b   [✓ Available]                     │
│   • qwen2.5:32b       [✗ Not pulled]                    │
│                                                          │
├─ Recommended Models by Task ─────────────────────────────┤
│                                                          │
│ Critical code  → Claude Opus 4.5                        │
│ Code review    → GPT-5.2 (preserves Claude)             │
│ Reasoning      → DeepSeek R1 70B (free, local)          │
│ Simple edits   → Gemini Flash                           │
│ Bulk work      → Llama 3.3 70B (free, local)            │
│                                                          │
└──────────────────────────────────────────────────────────┘

Config: ~/.model-router/config.yaml
Models: ~/.model-router/models.yaml

Commands:
  /model-router:configure  - Change settings
  /model-router:recommend  - Get recommendation for a task

4. Show Warnings

If any issues detected:

⚠ Warnings:
  • Codex CLI installed but not authenticated. Run: codex auth
  • Ollama running but no large models. Consider: ollama pull llama3.3:70b
  • Tracking disabled - Model Router can't learn your preferences

5. Show Recent Routing (if tracking enabled)

If tracking is enabled and ~/.model-router/tracking.jsonl exists:

Recent routing (last 7 days):
  • code_review: GPT-5.2 (8x), Claude Sonnet (3x)
  • reasoning: DeepSeek R1 (5x)
  • simple_edits: Gemini Flash (12x)

Run /model-router:learn to analyze patterns and improve recommendations.