|
|
--- |
|
|
title: Mcp4rdf |
|
|
emoji: 🔗 |
|
|
colorFrom: blue |
|
|
colorTo: purple |
|
|
sdk: gradio |
|
|
sdk_version: 5.33.0 |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
license: cc0-1.0 |
|
|
tag: "mcp-server-track" |
|
|
--- |
|
|
|
|
|
# Mcp4rdf - RDF Validation Server with AI |
|
|
|
|
|
**Space Link:** [https://huggingface.co/spaces/jimfhahn/mcp4rdf](https://huggingface.co/spaces/jimfhahn/mcp4rdf) |
|
|
|
|
|
This is a Gradio app that serves as an RDF validation tool and an MCP (Model Context Protocol) server. It allows users to validate RDF/XML content against SHACL schemas, and provides AI-powered suggestions and corrections for validation errors using a Hugging Face Inference Endpoint. |
|
|
|
|
|
**Track:** MCP Server / Tool |
|
|
**Tag:** `mcp-server-track` |
|
|
|
|
|
## MCP Server in Action |
|
|
|
|
|
The MCP server can be tested by connecting an MCP client to the Gradio app's API endpoint. |
|
|
The application is live and can be accessed at the Space Link above. |
|
|
|
|
|
## Features: |
|
|
- Real-time RDF/XML validation against SHACL schemas (Monograph and custom templates). |
|
|
- AI-powered error suggestions and corrections via a Hugging Face Inference Endpoint. |
|
|
- Built-in examples for valid, invalid, and BibFrame RDF. |
|
|
- Auto-validation as you type. |
|
|
- Easy-to-use Gradio interface. |
|
|
- Exposes MCP server functionality for clients like Claude Desktop. |
|
|
|
|
|
## MCP Server Tools Exposed: |
|
|
- `validate_rdf_tool`: Validate RDF/XML against SHACL shapes. |
|
|
- `get_ai_suggestions`: Get AI-powered fix suggestions. |
|
|
- `get_ai_correction`: Generate corrected RDF/XML. |
|
|
- `get_rdf_examples`: Retrieve example RDF snippets. |
|
|
|
|
|
## Setup and Deployment on Hugging Face Spaces: |
|
|
|
|
|
1. **Hugging Face Space:** This application is deployed at [https://huggingface.co/spaces/jimfhahn/mcp4rdf](https://huggingface.co/spaces/jimfhahn/mcp4rdf). |
|
|
2. **Inference Endpoint:** The application is configured to use a specific Hugging Face Inference Endpoint for AI features. |
|
|
3. **API Key:** To enable AI features, the `HF_API_KEY` (your Hugging Face API token with access to the Inference Endpoint) must be set as a Secret in the Space settings. |
|
|
* Go to your Space → Settings → Repository secrets. |
|
|
* Add: `HF_API_KEY` = `your_huggingface_api_key_here`. |
|
|
4. **Requirements:** The Space automatically installs dependencies from `requirements.txt`. |
|
|
|
|
|
## MCP Connection: |
|
|
1. When deployed on Hugging Face Spaces, the MCP server is available at: |
|
|
`https://jimfhahn-mcp4rdf.hf.space/gradio_api/mcp/sse` (Note: The Space ID `jimfhahn-mcp4rdf` might vary slightly based on HF naming conventions, please verify the exact URL from your Space). |
|
|
2. Use this URL in your MCP client's configuration (e.g., Claude Desktop). |
|
|
|
|
|
## Acknowledgments: |
|
|
This application utilizes concepts, code patterns, and profiles inspired by the work of the [Bibframe Interoperability Group (BIG) DCTAP](https://bf-interop.github.io/DCTap/) and [The bf-demo-validation-tool](https://github.com/bf-interop/bf-demo-validation-tool). |
|
|
|
