doris-mcp-server/examples/dify/dify_demo.md

Dify Example: Integrating Doris MCP Server

This document demonstrates how to integrate and use doris-mcp-server in Dify to perform Doris SQL calls via MCP.

Table of Contents

• Prerequisites
• Installation & Configuration
• Starting the MCP Server
• Ngrok Tunnel (Optional)
• Installing & Configuring the Plugin in Dify
• Creating a Dify App
• Adding MCP Tools
• Example Calls

---

Prerequisites

1. Clone the repository:

   git clone https://github.com/apache/doris-mcp-server.git
   
   cd doris-mcp-server
   

2. Install the latest Python package and copy the environment template:

   pip install mcp-doris-server
   
   cp .env.example .env
   
   vim .env
   

   Fill in your Doris FE connection settings in .env:

   DORIS_HOST=localhost
   DORIS_PORT=9030
   DORIS_USER=root
   DORIS_PASSWORD=
   DORIS_DATABASE=information_schema
   
   DORIS_FE_HTTP_PORT=8030
   
   # Optional: custom BE host list
   DORIS_BE_HOSTS=
   

Starting the MCP Server

Run the startup script:

./start_server.sh

If successful, you will see logs like this:

---

Ngrok Tunnel (Optional)

If your Dify deployment requires a public endpoint, use ngrok:

ngrok http http://localhost:3000

Use the generated public URL in subsequent steps.

---

Installing & Configuring the Plugin in Dify

1. In the Dify console, go to Plugin Marketplace, search for and install MCP‑SSE / StreamableHTTP:

2. After installation, click Configure and set the URL to your public or local address (e.g., https://abc123.ngrok.io/mcp):

   {
     "doris_mcp_server": {
       "transport": "streamable_http",
       "url": "https://<your-domain>/mcp"
     }
   }
   

   

3. Click Save. If configured correctly, you will see a green Authorized indicator:

   

---

Creating a Dify App

1. In the Dify console, click New App → Blank App.

2. Select Agent as the template, set App Name (e.g., Doris ChatBI).

---

Instructions & Tool Configuration

Instruction Block

Paste the following into the Instruction field:

<instruction>
Use MCP tools to complete tasks as much as possible. Carefully read the annotations, method names, and parameter descriptions of each tool. Please follow these steps:

1. Analyze the user's question and match the most appropriate tool.
2. Use tool names and parameters exactly as defined; do not invent new ones.
3. Pass parameters in the required JSON format.
4. When calling tools, use:
   {"mcp_sse_call_tool": {"tool_name": "<tool_name>", "arguments": "{}"}}
5. Output plain text only—no XML tags.

<input>
User question: user_query
</input>

<output>
Return tool results or a final answer, including analysis.
</output>
</instruction>

Adding MCP Tools

In the Tools pane, click Add twice to add two entries both named mcp_sse (they will inherit transport and URL from the plugin):

---

Example Calls

List Tables in Database

• User: What tables are in the database?
• Result: Dify will call the MCP tool to run SHOW TABLES and return the list.

Sales Trend Over Ten Years

• User: What has been the sales trend over the past ten years in the ssb database, and which year had the fastest growth?
• Result: The tool will execute the SQL, calculate growth rates, and return a trend chart or table.

<!--
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements.  See the NOTICE file
... (license omitted for brevity)
-->

# Dify Example: Integrating Doris MCP Server

This document demonstrates how to integrate and use the `doris-mcp-server` in Dify to perform Doris SQL calls via MCP.

## Table of Contents

- [Prerequisites](#prerequisites)
- [Installation & Configuration](#installation--configuration)
- [Starting the MCP Server](#starting-the-mcp-server)
- [Ngrok Tunnel (Optional)](#ngrok-tunnel-optional)
- [Installing & Configuring the Plugin](#installing--configuring-the-plugin)
- [Creating a Dify App](#creating-a-dify-app)
- [Instructions & Tool Configuration](#instructions--tool-configuration)
- [Example Calls](#example-calls)

---

## Prerequisites

1. Clone the repository:
   ```bash
   git clone https://github.com/apache/doris-mcp-server.git
   cd doris-mcp-server

2. Install the Python package and prepare the environment file:

   pip install mcp-doris-server
   cp .env.example .env
   vim .env
   

3. In .env, set your Doris front-end connection:

   DORIS_HOST=localhost
   DORIS_PORT=9030
   DORIS_USER=root
   DORIS_PASSWORD=
   DORIS_DATABASE=information_schema
   DORIS_FE_HTTP_PORT=8030
   # Optional: custom BE host list
   DORIS_BE_HOSTS=
   

---

Installation & Configuration

1. Install the MCP server package:

   pip install mcp-doris-server
   

2. Copy and edit the environment template as shown above.

---

Starting the MCP Server

Run the startup script:

./start_server.sh

If successful, you will see logs like this:

---

Ngrok Tunnel (Optional)

If your Dify deployment requires a public endpoint, use ngrok:

ngrok http http://localhost:3000

Use the generated URL in the plugin configuration.

---

Installing & Configuring the Plugin

1. In the Dify console, go to Plugin Marketplace, search for MCP‑SSE / StreamableHTTP, and install it.

2. Click Configure, and set the URL (e.g., https://<your-domain>/mcp):

{
  "doris_mcp_server": {
    "transport": "streamable_http",
    "url": "https://<your-domain>/mcp"
  }
}

3. Click Save. A green Authorized indicator should appear:

---

Creating a Dify App

1. In the Dify console, click New App → Blank App.

2. Select Agent as the template, set App Name (e.g., Doris ChatBI).

---

Instructions & Tool Configuration

Instruction Block

Paste the following into the Instruction field:

<instruction>
Use MCP tools to complete tasks as much as possible. Carefully read the annotations, method names, and parameter descriptions of each tool. Please follow these steps:

1. Analyze the user's question and match the most appropriate tool.
2. Use tool names and parameters exactly as defined; do not invent new ones.
3. Pass parameters in the required JSON format.
4. When calling tools, use:
   {"mcp_sse_call_tool": {"tool_name": "<tool_name>", "arguments": "{}"}}
5. Output plain text only—no XML tags.

<input>
User question: user_query
</input>

<output>
Return tool results or a final answer, including analysis.
</output>
</instruction>

Adding MCP Tools

In the Tools pane, click Add twice to add two entries both named mcp_sse (they will inherit transport and URL from the plugin):

---

Example Calls

List Tables in Database

• User: What tables are in the database?
• Result: Dify calls the MCP tool to run SHOW TABLES and returns the list.

Sales Trend Over Ten Years

• User: What has been the sales trend in the ssb database over the past ten years, and which year saw the fastest growth?
• Result: The tool executes SQL, computes growth rates, and returns a chart or table.