Moe Loubani
Back to projects
Libvips MCP
MCPSharpWASMTypeScript

Libvips MCP

300+ Image Operations for AI

An MCP server exposing the full power of libvips image processing to AI agents, with 300+ operations available.

Visit Libvips MCP

Overview

Libvips MCP gives AI agents access to professional-grade image processing through the Model Context Protocol. From simple resizes to complex operations like FFT analysis and morphological transforms, AI can now manipulate images with precision.

Technical Challenges

Libvips has hundreds of operations with complex parameter combinations. The challenges were:

  • Exposing operations in an AI-friendly way
  • Handling large images without memory issues
  • Supporting both Sharp (Node.js) and wasm-vips (browser)
  • Documenting operations for AI understanding

Solution

A dual-implementation approach:

  • Sharp for Node.js environments (fastest)
  • wasm-vips for browser/edge environments
  • Intelligent operation grouping for discoverability
  • Rich parameter descriptions for AI comprehension
  • Streaming support for large images

Key Features

  • Resize, crop, rotate, flip
  • Color space conversions
  • Compositing and blending
  • Filters and convolutions
  • FFT and frequency domain ops
  • Morphological operations
  • Format conversion (JPEG, PNG, WebP, AVIF)