Stats
Actions
Tags
From image-gen
Use when the user wants to edit, modify, transform, restyle, or create variations of an existing image using AI. Triggers on: edit this image, change the style, modify the photo, transform the picture, restyle, image variation, change the background, apply style transfer, make it look vintage, add text to image, remove background, change colors.
How this skill is triggered — by the user, by Claude, or both
Slash command
/image-gen:image-editThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Edit and transform existing images using FAL AI's image-to-image models. Supports style transfer,
Edit and transform existing images using FAL AI's image-to-image models. Supports style transfer, background changes, text modifications, targeted edits, and aesthetic transformations.
| Edit Type | Default Model | Endpoint |
|---|---|---|
| General edits | Seedream v4.5 | fal-ai/bytedance/seedream/v4.5/edit |
| Aesthetic transforms | Grok Imagine | xai/grok-imagine-image/edit |
| Fast edits | FLUX.2 Flash | fal-ai/flux-2/flash/edit |
| Precision edits | FLUX.2 [klein] | fal-ai/flux-2/klein/4b/edit |
| Text changes | Nano Banana Pro | fal-ai/nano-banana-pro/edit |
| Multi-reference | Seedream v4.5 | fal-ai/bytedance/seedream/v4.5/edit |
Read references/image-to-image.md for handling local files.
Determine what the user wants to change:
If unclear, ask the user what specific changes they want.
Read references/model-catalog.md for full details. Select based on edit type:
prompt-architect agent with the edit description and target modelguidance_scale based on desired change magnitude:
Read references/image-to-image.md for the full strength/guidance tuning guide.
Read references/prompt-engineering.md for edit-specific prompt strategies.
Execute the FAL API call:
./scripts/fal-submit.sh "{endpoint}" \
'{"prompt":"[edit description]","image_urls":["[source_url]"],"guidance_scale":[value]}' \
"./output/image-gen/edit-{model}-{timestamp}.png"
Read references/fal-auth.md to verify FAL_KEY is set.
edit- prefix in the filenamenpx claudepluginhub prakhar625/my-plugins --plugin image-genGuides AI-powered image editing with style transfer and object removal using Fal AI patterns. Useful for integrating image manipulation capabilities in apps.
Transforms or edits images using AI via the videoclaw CLI. Converts reference images to different styles (e.g., character grids) using providers like Volcengine or Gemini.
Provides prompting techniques for AI image generation and editing models on Replicate. Covers natural language prompts, photographic vocabulary, and iterative editing.