ai-image-video-studio
AI image/video: ComfyUI workflows and fixes (FLUX, SDXL, Wan, LoRA), product photos, image edits, video QA. ComfyUI workflow yap, node eksik, görseli düzenle.
SKILL.mdskills/claude/ai-image-video-studio/SKILL.md
GitHubAI Image & Video Studio
Choose the module
| Need | Module |
|---|---|
| New ComfyUI workflow or capability (FLUX, SDXL, SD3, Wan, Hunyuan, LTXV, Mochi, Cosmos, LoRA, ControlNet, IPAdapter, PuLID, inpaint, upscale, txt2vid, img2vid) | comfyui-workflow |
Existing ComfyUI JSON, runtime/startup error, missing nodes/models, Failed to fetch, VRAM/speed, identity drift, adding a module without breaking the main flow |
comfyui-workflow-guardian |
| AI product photos, packshots, lifestyle scenes, e-commerce image sets | product-photography |
| Edit an existing image while preserving identity and details | image-editing-workflow |
| Check a rendered video or export (resolution, fps, audio, subtitles, frames) | video-delivery-qa |
Banners, brand identity and social graphics → the separate ui-ux-pro-max family skills (banner-design, brand, design).
Shared rules
- Use the ComfyUI MCP or local API tools when connected; otherwise produce files the user can load.
- Start from a validated template; output valid JSON; list exact models and custom nodes with sources.
- Back up before editing a working workflow; prefer the smallest fix; never delete user models or outputs.
- Preserve faces, logos, text and product details exactly; never upload personal images to a new provider without consent.
Module map
Open only the module(s) the request needs and read the module file completely before acting. Several modules may combine in one task.
| Module | Use when | File |
|---|---|---|
comfyui-workflow |
Automatically design or substantially modify a ComfyUI workflow when the user wants a new image/video pipeline or capability for FLUX, SDXL, SD3, Wan, Hunyuan, LTXV, Moc… | MODULE.md |
comfyui-workflow-guardian |
Automatically audit, repair, refactor, and stabilize an existing ComfyUI workflow or Windows runtime when the user shares workflow JSON, an error/screenshot, or reports … | MODULE.md |
product-photography |
Plan, generate, edit, or review AI product photography for e-commerce, packaging, ads, catalogs, and product launches. | MODULE.md |
image-editing-workflow |
Apply a source-preserving workflow to plan, execute, and verify image edits with the active host's approved image-editing capability. | MODULE.md |
video-delivery-qa |
Inspect, validate, and prepare video deliverables for export or handoff using available local tools such as ffprobe, frame sampling, waveform/audio inspection, captions,… | MODULE.md |
Supporting files (open only when the module points to them):
comfyui-workflow-guardian: identity-modules.md, provenance-and-recovery.md, runtime-performance.md, runtime-profile.md, source-notes.md, workflow-editing-checklist.md; scripts:modules/comfyui-workflow-guardian/scripts/check_comfyui_workflow.py