mcp-iiif-images
Model Context Protocol (MCP) tools for IIIF image access
Documentation
MCP for IIIF Images
A Model Context Protocol (MCP) server for working with IIIF (International Image Interoperability Framework) manifests and images. See this video for a demonstration.
Features
This MCP server contains the following tools:
- `fetch_iiif_manifest`: Fetch a IIIF manifest from a URL. (Note that clients may have difficulty processing large amounts of JSON.)
- `fetch_iiif_image`: Retrieve a IIIF image from a base URI, fetching info.json and returning the image data (default: max 1500px dimension, max 800,000 pixels total)
- `fetch_iiif_image_region`: Retrieve a specific region of a IIIF image using percentage coordinates, with the region scaled to fit within the same constraints
*Caveats*
- The code scales the images to dimensions acceptable to Claude. It will not work with a Level 0 Image API implementation.
- Claude may not process some IIIF Manifests due to the size of the file.
Claude Desktop Configuration
Install from Claude Desktop Extension (DXT) file
1. Install and log in to Claude Desktop
2. Download the `.dxt` file from the latest release
3. Double-click the `.dxt` file
4. Install the extension when prompted by Claude
Install from Source Code
1. Clone this repository
2. Install dependencies: `npm install`
3. Make the server executable: `chmod +x server/server.js`
To use this MCP server with Claude Desktop, add the following configuration to your Claude Desktop config file. Adjust the file path as necessary. You may also need to provide the full path to your `node` command.
macOS
Edit `~/Library/Application Support/Claude/claude_desktop_config.json`:
{
"mcpServers": {
"mcp-iiif-images": {
"command": "node",
"args": ["/PATH/TO/mcp-iiif-images/server/server.js"]
}
}
}Windows
Edit `%APPDATA%\Claude\claude_desktop_config.json`:
{
"mcpServers": {
"mcp-iiif-images": {
"command": "node",
"args": ["C:\\path\\to\\mcp-iiif-images\\server\\server.js"]
}
}
}Linux
Edit `~/.config/Claude/claude_desktop_config.json`:
{
"mcpServers": {
"mcp-iiif-images": {
"command": "node",
"args": ["/path/to/mcp-iiif-images/server/server.js"]
}
}
}General MCP Usage
The server supports two transport modes:
Standard Mode (stdio)
Run the server using stdio transport (default):
npm start
# or
node server/server.jsHTTP Streaming Mode
Run the server using HTTP streaming transport:
node server/server.js --http
# or with custom port
node server/server.js --http --port 8080Command-line Options
- `--http`: Use HTTP streaming transport instead of stdio
- `--port PORT`: Port number for HTTP server (default: 3000)
- `--help`: Show help message
When using HTTP mode, the server will start an HTTP server with the following endpoints:
- `GET /sse`: Establish Server-Sent Events connection
- `POST /messages?sessionId=`: Send MCP messages
HTTP Streaming Mode
For HTTP streaming mode, you'll need to start the server manually with the `--http` flag and then configure Claude Desktop to connect via HTTP:
node server/server.js --http --port 3000Then configure Claude Desktop to use the HTTP transport (refer to Claude Desktop documentation for HTTP transport configuration).
Note: Update the path in the `args` array to match the actual location where you've installed this server.
After updating the configuration, restart Claude Desktop for the changes to take effect.
Testing
To run the tests:
# Run tests once
npm test
# Run tests in watch mode (automatically re-runs on file changes)
npm run test:watch
# Run tests with coverage
npm test -- --coverageThe project uses Vitest as the testing framework, which provides:
- Fast execution with ES modules support
- Jest-compatible API with better error messages
- Built-in coverage reporting
- Watch mode for development
Available Tools
fetch_iiif_manifest
Fetches and validates a IIIF manifest from a URL.
Parameters:
- `url` (required): The URL of the IIIF manifest to fetch
Example usage:
Please fetch the IIIF manifest from https://example.com/manifest.jsonfetch_iiif_image
Retrieve a IIIF image from a base URI, fetching info.json and returning the image data.
Parameters:
- `baseUri` (required): Base URI of the IIIF Image API resource (without /info.json)
Example usage:
Fetch the IIIF image at https://example.com/iiif/image123fetch_iiif_image_region
Retrieve a specific region of a IIIF image using percentage coordinates, with the region scaled to fit within the same constraints. Use this to fetch regions of interest at higher detail for more accurate image description and analysis.
Parameters:
- `baseUri` (required): Base URI of the IIIF Image API resource (without /info.json)
- `region` (required): Region in pct: format (e.g., 'pct:20,20,50,50' for x,y,width,height as percentages)
Example usage:
Fetch a region from the IIIF image at https://example.com/iiif/image123 with region pct:10,10,50,50Note that you can use these tools together in the course of a conversation, for example:
Fetch the IIIF image at https://example.com/iiif/image123 and describe it.
...
Zoom in on the text at the bottom of the page and transcribe it.Frequently asked questions
What is mcp-iiif-images?
mcp-iiif-images is Model Context Protocol (MCP) tools for IIIF image access
How do I install mcp-iiif-images?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is mcp-iiif-images open source?
Yes — it is hosted on GitHub at https://github.com/mikeapp/mcp-iiif-images and has 6 stars.
Related MCP tools
🔥 Official Firecrawl MCP Server - Adds powerful web scraping and search to Cursor, Claude and any other LLM clients.
Use any LLMs (Large Language Models) for Deep Research. Support SSE API and MCP server.
Enhanced MCP server for interactive user feedback and command execution in AI-assisted development, featuring dual interface support (Web UI and Desktop Application) with intelligent environment detection and cross-platform compatibility.
A powerful Zotero AI and MCP plugin with ChatGPT, Gemini 3.7, Claude Fable 5, Claude Opus 5, DeepSeek V4, Grok, OpenRouter, Kimi k3, GLM 5.3, SiliconFlow, GPT-oss, Gemma 4, Qwen 3.8
Connect your browser to AI models. Just use Dia on Chrome, Arc or Firefox.
文颜 MCP Server 可以让 AI 自动将 Markdown 文章排版后发布至微信公众号。
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP