MCKRUZ/ComfyUI-Expert
View on GitHub9 skills discovered in this repository.
38$npx skills add MCKRUZ/ComfyUI-ExpertSkills in this Repository
Generate valid ComfyUI workflow JSON from natural language. Handles txt2img, img2img, ControlNet, identity, video, upscaling, inpainting.
Connect to ComfyUI REST API to queue image generation workflows, monitor progress, and retrieve outputs with validation.
Conduct guided interviews to extract creative vision and synthesize model-appropriate image prompts with settings recommendations.
Systematically diagnose and fix ComfyUI errors, workflow failures, and quality issues using error patterns and community workarounds.
Guide to preparing datasets, configuring LoRA training for FLUX and SDXL, and evaluating checkpoints for character consistency in image generation.
Track ComfyUI models, nodes, and techniques across YouTube, GitHub, and HuggingFace. Extract timestamped findings and flag stale information.
Combine clips, audio, and assets into final videos using FFmpeg or Remotion. Handles mixing, normalization, subtitles, and format conversion.
Create and maintain video production projects with character profiles, workflow histories, and asset tracking in ComfyUI.
Orchestrate video publishing to YouTube, Shorts, Instagram Reels, and TikTok with optimized metadata generation and pre-publish checklists.