troubleshooting

Diagnose ComfyUI errors with debugging commands and targeted fixes.

530|85|Updated Feb 15, 2026
One-click install
npx skills add https://github.com/artokun/comfyui-mcp --skill troubleshooting-artokun
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: troubleshooting
Source: https://github.com/artokun/comfyui-mcp/tree/main/plugin/skills/troubleshooting
Command: npx skills add https://github.com/artokun/comfyui-mcp --skill troubleshooting-artokun

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps users diagnose and resolve common errors encountered when using ComfyUI, such as out-of-memory issues, missing nodes, and device mismatches, ensuring smoother workflow execution.

Core Features & Use Cases

  • Error Diagnosis: Provides systematic steps to identify the root cause of ComfyUI failures.
  • Specific Error Solutions: Offers targeted fixes for common problems like OOM errors, device mismatches, missing nodes, NaN tensors, and black images.
  • Debugging Commands: Includes direct commands to retrieve logs, check system stats, and verify node/model availability.
  • Use Case: If your ComfyUI workflow suddenly starts producing black images without any error message, this Skill can guide you through checking sampler settings, VAE compatibility, and prompt configurations to find the solution.

Quick Start

Use the troubleshooting skill to diagnose why my ComfyUI workflow is failing with a CUDA out of memory error.

Frequently Asked Questions about troubleshooting

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I fix a CUDA out of memory error in ComfyUI?

To fix a ComfyUI CUDA out of memory error, you can follow systematic diagnostic steps to check VRAM usage, adjust batch sizes, and verify model loading configurations to free up GPU memory.

Why does my ComfyUI workflow produce black images without any error message?

ComfyUI workflows produce black images when there are VAE compatibility issues, incorrect sampler settings, or prompt configuration errors, requiring systematic checks of these specific components to resolve.

What is the best way to resolve device mismatch errors in ComfyUI?

Resolving device mismatch errors in ComfyUI requires verifying CUDA and Torch versions, checking system stats, and ensuring model loading targets the correct GPU device using specific debugging commands.

How do I troubleshoot NaN tensors and missing nodes in ComfyUI?

Troubleshooting NaN tensors and missing nodes in ComfyUI involves running direct debugging commands to retrieve logs, verify node availability, and apply targeted fixes for model loading failures.

Does ComfyUI Desktop handle error resolution differently than the CLI version?

ComfyUI Desktop and CLI versions have differences in error resolution and model loading, requiring specific diagnostic strategies to compare system logs and verify environment configurations for each platform.