content-replace-inpaint

name: Content Replace & Inpaint description: Replace, edit, or regenerate specific regions of images using AI-powered content replacement version: 1.0.0 allowed-tools: - flux_fill - flux_kontext - search_replace - inpaint - erase

Content Replace & Inpaint

Intelligently replace, edit, or regenerate specific regions of images using state-of-the-art AI models.

Overview

This skill provides powerful content replacement capabilities using both mask-based inpainting and semantic search-based replacement. Perfect for removing unwanted objects, replacing backgrounds, or editing specific elements while maintaining context.

Available Tools

FLUX-Based (Black Forest Labs)

• {{flux_fill}}: High-quality masked region filling - specify what to generate in masked areas
• {{flux_kontext}}: Context-aware editing - modify images based on text instructions

Stability AI

• {{search_replace}}: Semantic search and replace - find objects by description and replace them
• {{inpaint}}: Fill masked regions with generated content
• {{erase}}: Remove unwanted objects seamlessly

Tool Comparison

Usage Examples

Example 1: Replace Sky with Sunset (Mask-based)

Use {{flux_fill}} with an image and a mask (white areas indicate regions to fill):

{
  "prompt": "dramatic orange and purple sunset sky, golden hour lighting, beautiful clouds",
  "guidance": 3.5,
  "prompt_upsampling": true,
  "output_format": "png"
}

Blob inputs: image (source photo), mask (sky region marked white)

Example 2: Change Object by Description

Use {{search_replace}} to find and replace objects semantically:

{
  "search_prompt": "red car in parking lot",
  "prompt": "blue sports car, matching perspective and lighting",
  "output_format": "png"
}

Blob inputs: image (source photo)

Example 3: Context-Aware Edit

Use {{flux_kontext}} for holistic changes without a mask:

{
  "prompt": "Change the person's outfit to a red dress while keeping everything else the same",
  "guidance": 3.5,
  "prompt_upsampling": true
}

Blob inputs: image (source photo)

Example 4: Remove Unwanted Objects

Use {{erase}} with a mask to seamlessly remove objects:

{
  "output_format": "png"
}

Blob inputs: image (source photo), mask (object region marked white)

Example 5: Fill with Content

Use {{inpaint}} to fill masked regions with new content:

{
  "prompt": "beautiful flower garden with roses and tulips",
  "negative_prompt": "blurry, low quality",
  "output_format": "png"
}

Blob inputs: image (source photo), mask (region to fill marked white)

Workflow Guide

For Precise Edits (Mask Required)

1. Prepare the mask: Create a black and white image where white indicates areas to edit
2. Choose your tool:
   • {{flux_fill}} for highest quality replacement
   • {{inpaint}} for quick fills
   • {{erase}} for object removal
3. Provide context: Include details about surrounding areas in your prompt

For Semantic Edits (No Mask)

1. Describe what to find: Use {{search_replace}} with a clear search prompt
2. Describe replacement: Specify what should replace it
3. Consider context: Include style and lighting cues for seamless blending

For Holistic Edits

1. Use {{flux_kontext}}: Describe the desired change in natural language
2. Be specific: "Change the weather to rainy" or "Make the room look modern"
3. Preserve elements: The AI understands context and preserves unrelated elements

Parameters

{{flux_fill}}

{{flux_kontext}}

{{search_replace}}

{{inpaint}} / {{erase}}

Best Practices

Creating Good Masks

• Use pure white (#FFFFFF) for areas to edit
• Use pure black (#000000) for areas to preserve
• Add slight feathering (2-5px) for smoother blending
• Ensure mask dimensions match the source image

Writing Effective Prompts

For {{flux_fill}}:

✅ "A sunset sky with orange and purple clouds, golden hour lighting matching the scene"
❌ "sunset"

For {{search_replace}}:

✅ Search: "red car in parking lot"
✅ Replace: "blue sports car, matching perspective and lighting"
❌ Search: "car" → Too generic

For {{flux_kontext}}:

✅ "Change the person's hair color to blonde while keeping everything else the same"
❌ "blonde hair" → Too vague

Maintaining Consistency

1. Match lighting and perspective in replacement prompts
2. Reference surrounding elements for coherent results
3. Use lower guidance (2-3) for subtle changes
4. Use higher guidance (4-5) for dramatic transformations

Common Use Cases

Limitations

• Mask quality significantly affects results
• Complex semantic searches may have unpredictable results
• Large replaced areas may show inconsistencies
• Processing time: 15-60 seconds depending on complexity