Dify API MCP

A Model Context Protocol (MCP) server that integrates with Dify API to upload files to knowledge bases.

Features

• Upload local files to Dify knowledge base via MCP
• Configurable segmentation settings
• Built-in error handling and validation
• Easy integration with Claude Desktop

Installation

npm install dify-api-mcp

Configuration

Add the MCP server to your Claude Desktop configuration file:

macOS

~/Library/Application Support/Claude/claude_desktop_config.json

Windows

%APPDATA%\Claude\claude_desktop_config.json

Configuration Example

{
  "mcpServers": {
    "dify-api-mcp": {
      "command": "node",
      "args": ["./node_modules/dify-api-mcp/src/server.js"],
      "env": {
        "DIFY_API_KEY": "your-api-key-here",
        "DIFY_DATASET_ID": "your-dataset-id-here",
        "DIFY_BASE_URL": "https://api.dify.ai/v1",
        "DIFY_SEPARATOR": "\\n",
        "DIFY_MAX_TOKENS": "1000"
      }
    }
  }
}

Environment Variables

• DIFY_API_KEY (required): Your Dify API key
• DIFY_DATASET_ID (required): Target dataset/knowledge base ID
• DIFY_BASE_URL (optional): Dify API base URL (defaults to https://api.dify.ai/v1)
• DIFY_SEPARATOR (optional): Segmentation separator (defaults to "\n")
• DIFY_MAX_TOKENS (optional): Maximum tokens per segment (defaults to 1000)

Segmentation Settings

You can configure document segmentation using these environment variables:

• DIFY_SEPARATOR: Text separator for splitting documents (default: "\n")
• DIFY_MAX_TOKENS: Maximum tokens per document segment (default: 1000)

Usage

After configuration, restart Claude Desktop. You can then use the following MCP tool:

createByFile

Upload a local file to the Dify knowledge base.

Parameters:

• filePath (string, required): Path to the local file to upload

Example:

Please upload the file /path/to/document.pdf to the knowledge base

API Configuration

The MCP server uses the following fixed settings for Dify API calls:

• indexing_technique: "high_quality"
• doc_form: "text_model"
• process_rule.mode: "automatic"

Error Handling

The server includes comprehensive error handling for:

• File validation (existence, size limits)
• Network connectivity issues
• API authentication and authorization
• Dify API errors

Development

# Install dependencies
npm install

# Run the server
npm start

License

MIT