tensors/TODO.md

TODO: ComfyUI CLI & API Integration

Phase 1: Core Client (tensors/comfyui.py)

• Step 1.1: Create ComfyUI client module with basic query functions
  • get_system_stats() - System stats (GPU, RAM)
  • get_queue_status() - Queue status
  • get_loaded_models() - List loaded checkpoints/loras
  • get_history() - View history
  • Also: clear_queue(), get_object_info(), get_image()
• Step 1.2: Add workflow execution with WebSocket progress tracking
  • queue_prompt() - Queue a workflow
  • run_workflow() - Run workflow with progress callback (uses polling)
• Step 1.3: Add simple text-to-image generation
  • generate_image() - Text-to-image with embedded workflow template
  • Include SDXL/Flux-compatible default workflow (DEFAULT_WORKFLOW_TEMPLATE)

Phase 2: CLI Commands (tensors/cli.py)

• Step 2.1: Add comfy subcommand group with status commands
  • tsr comfy status - System stats (GPU, RAM, queue)
  • tsr comfy queue - Queue status
  • tsr comfy queue --clear - Clear queue
  • tsr comfy models - List loaded models
  • tsr comfy history [PROMPT_ID] - View history
• Step 2.2: Add generation commands
  • tsr comfy generate "prompt" - Simple text-to-image with full options
  • tsr comfy run workflow.json - Run arbitrary workflow
  • Rich progress output via console

Phase 3: Server API Routes (tensors/server/comfyui_api_routes.py)

• Step 3.1: Create new router with query endpoints
  • GET /api/comfyui/status - System stats
  • GET /api/comfyui/queue - Queue status
  • DELETE /api/comfyui/queue - Clear queue
  • GET /api/comfyui/models - List loaded models
  • GET /api/comfyui/history - List history
  • GET /api/comfyui/history/{prompt_id} - Get specific result
• Step 3.2: Add generation endpoints
  • POST /api/comfyui/generate - Text-to-image generation
  • POST /api/comfyui/workflow - Run arbitrary workflow
• Step 3.3: Register router in server/__init__.py (with API key auth)

Phase 4: Configuration (tensors/config.py)

• Step 4.1: Add ComfyUI config functions
  • get_comfyui_url() - Get ComfyUI backend URL (COMFYUI_URL env -> config -> default)
  • get_comfyui_defaults() - Get default generation settings from config
  • Environment variable: COMFYUI_URL
  • Config section: [comfyui] with url, default_model, width, height, steps, cfg, sampler, scheduler
  • Updated comfyui.py to use config functions

---

Architecture Reference

tensors/
├── comfyui.py              # Core client (NEW) - shared by CLI and server
├── cli.py                  # Add `tsr comfy` subcommands
└── server/
    ├── comfyui_routes.py       # Existing proxy (unchanged)
    └── comfyui_api_routes.py   # New programmatic API routes (NEW)

CLI Commands

tsr comfy status              # System stats (GPU, RAM, queue)
tsr comfy queue               # Current queue status
tsr comfy queue --clear       # Clear queue
tsr comfy models              # List loaded checkpoints/loras
tsr comfy history [PROMPT_ID] # View history or specific result
tsr comfy generate "prompt"   # Simple text-to-image
tsr comfy run workflow.json   # Run arbitrary workflow

Generate Options

tsr comfy generate "a cat" \
  -n "blurry, bad" \           # negative prompt
  -m "flux1-dev-fp8.safetensors" \  # model
  -W 1024 -H 1024 \            # dimensions
  --steps 20 --cfg 7.0 \       # sampling
  --seed 42 \                  # reproducibility
  -o ./output.png              # output path