Nutrient DWS

You can empower AI agents to create, edit, convert, redact, sign, and OCR documents by connecting to the Nutrient Document Web Service Processor API through this MCP server. It exposes natural language enabled document processing capabilities to your AI workflows, enabling seamless, production-grade document transformations.

How to use

To use this MCP server, connect your AI agent or client to the MCP endpoint configured in your environment. Once connected, you can instruct it in natural language to perform a wide range of document operations such as merging PDFs, redacting sensitive data, applying digital signatures, converting formats, and extracting structured data. Your AI can issue commands like merging multiple documents, redacting personal information, signing contracts with visible signatures, converting PDFs to other formats, or performing OCR on scanned documents in a chosen language.

How to install

Prerequisites: ensure Node.js 18 or newer is installed on your system. You also need an API key from Nutrient to access the Data Web Service Processor.

Install and run the MCP server using the standard runtime command shown in the setup examples.

# Install dependencies and start the MCP server locally via NPX
npx -y @nutrient-sdk/dws-mcp-server

Configuration and running notes

Configure your MCP client to connect to the server. You typically provide the API key and a sandbox path where input and output documents will be stored.

{
  "mcpServers": {
    "nutrient_dws": {
      "command": "npx",
      "args": ["-y", "@nutrient-sdk/dws-mcp-server"],
      "env": {
        "NUTRIENT_DWS_API_KEY": "YOUR_API_KEY_HERE",
        "SANDBOX_PATH": "/your/sandbox/directory"
      }
    }
  }
}

Sandbox mode and output location

Sandbox mode restricts file operations to a designated directory to improve security. Set the SANDBOX_PATH environment variable to enable it. Processed files are saved within the sandbox, and relative paths are resolved against the sandbox directory.

To guide where results are stored, specify output paths in natural language or create an output directory inside the sandbox.

Troubleshooting

If the MCP server does not appear in your IDE or client, verify Node.js version, confirm the config path is correct for your platform, and restart the client fully.

If you encounter API key errors, double-check the key at your Nutrient dashboard and ensure it is set correctly in the environment configuration.

Security and data handling

The server supports redaction, password protection, and access controls for sensitive documents. Use sandbox mode to limit file operations to a safe directory and to prevent leakage of files outside the sandbox.

Supported tools and capabilities

Document processing tools let you merge, convert, OCR, redact, and manipulate documents. You can also digitally sign PDFs, browse sandbox contents, and examine directory structures.

Available tools

document_processor

All-in-one document processing: merge PDFs, convert formats, apply OCR, watermark, rotate, redact, flatten annotations, extract text/tables/key-value pairs, and more

document_signer

Digitally sign PDFs with PAdES-compliant CMS or CAdES signatures, with customizable visible or invisible signature appearances

sandbox_file_tree

Browse files in the sandbox directory when sandbox mode is enabled

directory_tree

Browse directory contents when sandbox mode is disabled