Orion Vision

Provides integration with Azure Form Recognizer to analyze documents and extract form data via MCP.
  • typescript

2

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "cognitive-stack-orion-vision-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "orion-vision-mcp@latest"
      ],
      "env": {
        "AZURE_FORM_RECOGNIZER_KEY": "your-key-here",
        "AZURE_FORM_RECOGNIZER_ENDPOINT": "your-endpoint-here"
      }
    }
  }
}

Orion Vision MCP Server lets you connect Azure Form Recognizer capabilities to MCP clients so you can analyze documents and extract form data securely. It provides a smooth, type-safe integration that works with multiple MCP clients, enabling you to automate document processing tasks across receipts, invoices, IDs, and more.

How to use

You use Orion Vision MCP Server by running it as a local command (stdio) or by integrating it into your MCP client ecosystem. Once the server is running, your MCP clients can send document analysis requests to Azure Form Recognizer through the MCP interface. Typical tasks include analyzing documents, extracting form data from forms like invoices, and processing ID documents to pull structured information.

How to install

Prerequisites you need before installation:

  • Azure Form Recognizer / Document Intelligence endpoint and key
  • Node.js (v20 or higher)
  • Git (only needed if using Git installation method)
  • A MCP client you plan to use (e.g., Claude Desktop, Cursor)

Install the server using NPX provided by the MCP ecosystem. Run the following command in your terminal:

npx -y orion-vision-mcp@latest

If you prefer automatic installation for a specific client via Smithery, run this command to install Orion Vision MCP Server for Claude Desktop automatically via Smithery:

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

Additional configuration and usage notes

Configure your MCP clients to connect to the Orion Vision MCP Server by providing the Azure Form Recognizer credentials and pointing to the server command. Below are the example configurations shown for common clients.

Configure Cline to add Orion Vision MCP Server support. You can install it with a single click in the marketplace or add it manually in the MCP settings file. The manual setup example shows how to define the server in the Cline settings.

{
  "mcpServers": {
    "orion-vision-mcp": {
      "command": "npx",
      "args": ["-y", "orion-vision-mcp@latest"],
      "env": {
        "AZURE_FORM_RECOGNIZER_ENDPOINT": "your-endpoint-here",
        "AZURE_FORM_RECOGNIZER_KEY": "your-key-here"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

Usage in Claude Desktop App

After installation, fully close and re-open the Claude Desktop app to load the Orion Vision MCP Server. You should see a hammer icon in the bottom left, indicating MCP tools are available.

Common usage examples you can run once the server is connected include analyzing a document, extracting form data from an invoice, and processing an ID document to extract relevant information.

Troubleshooting

If you encounter issues, verify the following:

Server not found: ensure Node is installed, verify the MCP server command, and confirm the environment variables are correctly configured.

Azure Form Recognizer credentials: double-check the endpoint and key, ensure there are no extra spaces or quotes, and confirm they are set in the server configuration.

Document processing: verify the document URL is accessible, ensure the document type is supported, and confirm the document is not corrupted.

Available tools

Analyze a Document

Analyze a document using Azure Form Recognizer to obtain a structured understanding of its content.

Extract Form Data

Extract specific data fields from forms such as invoices to populate structured records.

Process ID Document

Process ID documents to extract identifying information and relevant attributes.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational