Orion Vision (Azure Form Recognizer)

🚀 Retinal Vision MCP Server

The Retinal Vision MCP Server enables seamless interaction between AI systems and various data sources and tools through the Model Context Protocol (MCP).

The Model Context Protocol (MCP) is an open standard that allows AI systems to interact seamlessly and connect to various data sources and tools, enabling secure two-way connections.

The Retinal Vision MCP Server offers the following features:

• Seamless integration with Azure Form Recognizer / Document Intelligence Service
• Document analysis and form data extraction capabilities
• Support for multiple document types (receipts, invoices, IDs, etc.)
• Type-safe operations using TypeScript

🚀 Quick Start

Prerequisites 🔧

Before you begin, make sure you have the following:

• Endpoint and key for Azure Form Recognizer / Document Intelligence Service
• Claude Desktop or Cursor
• Node.js (v20 or higher)
• Git installed (only required if using the Git installation method)

📦 Installation

Run with NPX

npx -y orion-vision-mcp@latest

Install via Smithery

To automatically install the Retinal Vision MCP Server for Claude Desktop via Smithery:

npx -y @smithery/cli install @orion-vision/mcp --client claude

⚙️ Configure MCP Clients

Configure Cline 🤖

The easiest way to set up the Retinal Vision MCP Server with Cline in VS Code is to click and install it from the marketplace:

1. Open Cline in VS Code.
2. Click the Cline icon in the left sidebar.
3. Go to the "MCP Server" tab (four squares).
4. Search for "Retinal Vision" and click "Install".
5. When prompted, enter your Azure Form Recognizer credentials.

Alternatively, you can manually copy the following to your configuration file:

{
  "mcpServers": {
    "orion-vision-mcp": {
      "command": "npx orion-vision-mcp",
      "args": []
    }
  }
}

Configure Cursor 🛏️

For users of Cursor, you can add the Retinal Vision MCP Server as a new language support:

1. Open your project.
2. Go to the "File" menu and select "Configure".
3. Add a new language support in the settings:

{
  "languageSupport": {
    "orion-vision-mcp": {
      "path": "npx orion-vision-mcp"
    }
  }
}

💻 Usage Examples

Basic Usage

1. Analyze a document:

Analyze the document at "https://example.com/document.pdf" using Azure Form Recognizer.

2. Extract form data:

Extract data from the invoice at "https://example.com/invoice.pdf".

3. Process an ID document:

Process the ID document at "https://example.com/id.pdf" and extract relevant information.

🛠️ Troubleshooting

Common Issues

1. Server not found

   • Verify that the npm installation is correct by running npm --version.
   • Check the Claude Desktop configuration syntax.
   • Ensure that Node.js is installed correctly by running node --version.

2. Azure Form Recognizer credential issues

   • Confirm that your Azure Form Recognizer endpoint and key are valid.
   • Check that the credentials are correctly set in the configuration.
   • Make sure there are no extra spaces or quotes.

3. Document processing issues

   • Verify that the document URL is accessible.
   • Check that the document format is supported.
   • Ensure that the document is not corrupted or encrypted.

📄 Acknowledgments

• The Model Context Protocol (MCP) for providing the specification.
• Anthropic for providing Claude Desktop.
• Microsoft Azure for providing the Form Recognizer / Document Intelligence Service.