Image-Refinery (Rust)

A Model Context Protocol (MCP) server for image processing and metadata reconstruction, built with Rust.

[English] | 中文版

Features

• Geometric Transformations: Resizing, cropping, rotation, and flipping.
• Smart Crop: Content-aware cropping that automatically preserves visual focal points.
• Visual Filters: Blur, sharpen, grayscale, brightness/contrast, and rounded corners.
• Format Transcoding: High-quality conversion between PNG, JPG, WebP, and AVIF.
• Asset Optimization: Reduces file size through intelligent lossy and lossless compression.
• Layer Compositing: Overlays images or text with precise positioning and blend modes.
• Background Removal: Generates transparent PNG/WebP for logos, icons, and product images.
• Icon Generation: Creates App icon sets and multi-layer ICO files.
• Metadata Removal: Purges EXIF, GPS, and other metadata through pixel-level re-encoding.
• Metadata Editing: Sets custom EXIF tags (Copyright, Artist, Description, etc.) on images.
• Robust Format Detection: Automatically handles mismatched file extensions (e.g., JPEG content in a .png file).

Supported Environments

Compatible with MCP-enabled tools including:

• Antigravity, Windsurf, Cursor, Zed
• Roo Code, Cline, Augment Code
• Claude Desktop

Tool Reference

1. analyze_image

Extracts technical metadata from image files.

• Functions: Detects dimensions (width/height), file format, color space, and EXIF data (if available).
• Inputs: image_path (supports glob patterns like *.jpg).
• Output: JSON object with image specifications and EXIF metadata.

2. transform_image

Applies geometric transformations.

• Functions: Resizing, cropping, rotation, and flipping.
• Options: Supports smart_crop (content-aware) and manual crop regions.
• Parameters: image_path (input), width, height, rotate, flip, flop, quality (1-100).

3. resize_image

Specialized tool for image scaling.

• Functions: Adjusts dimensions while maintaining aspect ratio if only one dimension is provided.
• Parameters: image_path (input), width, height, quality (1-100).

4. stylize_image

Applies visual filters and adjustments.

• Functions: Blur, sharpen, grayscale conversion, brightness/contrast adjustment, and rounded corners.
• Parameters: image_path (input), blur, sharpen, grayscale, brightness, contrast, rounded_corners, quality.

5. convert_format

Transcodes images between different formats.

• Functions: Converts between PNG, JPG, WebP, and AVIF.
• Parameters: image_path (input), format (target), quality (1-100).
• Result: New file with the specified extension and no original metadata.

6. optimize_image

Reduces file size through compression.

• Functions: Minimizes storage footprint without changing dimensions.
• Parameters: image_path (input), quality (1-100) or lossless (boolean).

7. composite_image

Overlays images or text onto a base image.

• Functions: Adding watermarks, logos, or captions with precise positioning.
• Options: image_path (base), overlays array (path/text, x, y, font_size, color).

8. remove_background

Generates transparent background images.

• Functions: Removes background to extract subjects, outputting PNG or WebP with transparency.
• Parameters: image_path (input), force_color (hex), sensitivity (low/medium/high), format ("png" | "webp").

9. generate_icons

Generates sets of icons for application development.

• Functions: Creates a series of PNG icons at various sizes and an optional multi-layer .ico file.
• Parameters: image_path (input), sizes (array), generate_ico (boolean).

10. set_metadata

Sets image metadata (EXIF) by writing specified tags to the file.

• Mode: Overwrite-only. Existing metadata is replaced, not merged.
• Supported Tags: Copyright, Artist, Software, Description, UserComment, Make, Model.
• Supported Formats: JPG, PNG, WebP.
• Parameters: image_path, tags (key-value object).

Configuration

Add this entry to your MCP configuration file:

{
  "mcpServers": {
    "image-refinery": {
      "command": "npx",
      "args": ["-y", "@datafrog-io/image-refinery"],
      "env": {
        "IMAGE_REFINERY_API_KEY": "YOUR_LICENSE_KEY"
      }
    }
  }
}

© 2025 Datafrog LLC. All rights reserved.