mckruz avatar

Comfyui Troubleshooter

When your ComfyUI workflow throws a cryptic error or generates burned faces at 1024x1024, this skill walks you through systematic diagnosis. It classifies issues into server, workflow, quality, or performance buckets, then matches against a database of known patterns like CUDA OOM fixes, missing custom nodes, or InstantID overexposure. The quick reference covers the top 10 errors with actual solutions, not guesses. Honestly most useful when you're staring at "Expected scalar type BFloat16 but found Float" at 2am and just need someone to tell you to add the force-fp16 flag. It also handles the tedious stuff like mapping node types to their source packages and telling you exactly where to download missing models.

skills.sh installs
1.1K
GitHub stars
94
last updated
Jul 14, 2026

Install

$ npx -y skills add mckruz/comfyui-expert --skill comfyui-troubleshooter --agent claude-code

Installs into .claude/skills of the current project.

Files

ComfyUI Troubleshooter

Diagnoses and resolves ComfyUI issues across four categories: server errors, workflow errors, quality issues, and performance problems.

Diagnosis Process

Step 1: Classify the Error

CategorySymptomsFirst Check
ServerConnection refused, timeouts, crashesIs ComfyUI running? Check /system_stats
WorkflowNode errors, missing inputs, type mismatchesValidate workflow against inventory
QualityArtifacts, wrong identity, blurry outputCheck settings (CFG, weights, resolution)
PerformanceOOM, slow generation, VRAM errorsCheck VRAM usage, model sizes

Step 2: Gather Context

Collect before diagnosing:

  1. Error message (exact text)
  2. Workflow being executed (or description)
  3. Models involved (checkpoint, LoRA, ControlNet, etc.)
  4. Settings (CFG, steps, resolution, sampler)
  5. Hardware (from foundation/hardware-profile.md)
  6. Inventory (from state/inventory.json)

Step 3: Match Error Pattern

See references/troubleshooting.md for the full error database.

Quick Fix Reference

Top 10 Most Common Errors

Sponsors

Details

Repository
github.com/mckruz/comfyui-expert
Category
Generative Media
First seen
Jun 3, 2026

More from mckruz/comfyui-expert

All 10
1vibe-motion avatarremotion-3d-tickervibe-motion/skillsCreates infinite 3D vertical scrolling ticker animations in Remotion. Use when you need to build a parallax gallery, infinite image scroll, multi-column continu970skills.sh installs
2heygen-com avatarfaceswapheygen-com/skillsSwap faces in a video using AI via the HeyGen API. Use when: (1) Replacing a face in a video with another face, (2) Face swapping from a source image onto a tar903skills.sh installs
3black-forest-labs avatarflux-best-practicesblack-forest-labs/skillsComprehensive guide for BFL FLUX image generation models. Covers prompting, T2I, I2I, structured JSON, hex colors, typography, multi-reference editing, and mode854skills.sh installs
4bryanwhl avatarffmpeg-video-editorbryanwhl/ffmpeg-video-editorA complete FFmpeg-powered video editing agent. Use this skill whenever the user wants to do anything with video or audio files — trimming, cutting, transcoding,828skills.sh installs
5inference-sh avatarelevenlabs-musicinference-sh/skillsElevenLabs AI music generation - create original music from text prompts via inference.sh CLI. Capabilities: text-to-music, custom duration up to 10 minutes, ge821skills.sh installs
6inference-sh avatarelevenlabs-sound-effectsinference-sh/skillsGenerate AI sound effects from text descriptions with ElevenLabs via inference.sh CLI. Capabilities: text-to-sound-effect, custom duration, royalty-free audio. 783skills.sh installs