Files

Christian Tzolov 234d3ff919 refactor(mcp): update to use Spring Boot autoconfiguration and rename tool callbacks

Update MCP examples to use Spring Boot autoconfiguration instead of manual configuration.
Replace deprecated defaultTools method with defaultToolCallbacks across all examples.
Update documentation to reflect these changes and provide more detailed configuration examples.

- Update API from defaultTools to defaultToolCallbacks in all MCP examples
- Update README files with more detailed configuration examples

Signed-off-by: Christian Tzolov <christian.tzolov@broadcom.com>

2025-05-02 14:28:22 +03:00

3.3 KiB

Raw Permalink Blame History

Spring AI - Model Context Protocol (MCP) Brave Search Example

This example demonstrates how to create a Spring AI Model Context Protocol (MCP) client that communicates with the Brave Search MCP Server. The application shows how to build an MCP client that enables natural language interactions with Brave Search, allowing you to perform internet searches through a conversational interface. This example uses Spring Boot autoconfiguration to set up the MCP client through configuration files.

When run, the application demonstrates the MCP client's capabilities by asking a specific question: "Does Spring AI supports the Model Context Protocol? Please provide some references." The MCP client uses Brave Search to find relevant information and returns a comprehensive answer. After providing the response, the application exits.

Prerequisites

Java 17 or higher
Maven 3.6+
npx package manager
Git
OpenAI API key
Brave Search API key (Get one at https://brave.com/search/api/)

Setup

Install npx (Node Package eXecute): First, make sure to install npm and then run:
```
npm install -g npx
```

Clone the repository:

git clone https://github.com/spring-projects/spring-ai-examples.git
cd model-context-protocol/brave

Set up your API keys:

export OPENAI_API_KEY='your-openai-api-key-here'
export BRAVE_API_KEY='your-brave-api-key-here'

Build the application:
```
./mvnw clean install
```

Running the Application

Run the application using Maven:

./mvnw spring-boot:run

The application will execute a single query asking about Spring AI's support for the Model Context Protocol. It uses the Brave Search MCP server to search the internet for relevant information, processes the results through the MCP client, and provides a detailed response before exiting.

How it Works

The application integrates Spring AI with the Brave Search MCP server through Spring Boot autoconfiguration:

MCP Client Configuration

The MCP client is configured using configuration files:

application.properties:

spring.ai.mcp.client.stdio.servers-configuration=classpath:/mcp-servers-config.json

mcp-servers-config.json:

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-brave-search"
      ],
      "env": {
      }
    }
  }
}

This configuration:

Uses the Brave Search MCP server via npx
The Brave API key is passed from environment variables
Initializes a synchronous connection to the server

Chat Integration

The ChatClient is configured with the MCP tool callbacks in the Application class:

var chatClient = chatClientBuilder
        .defaultToolCallbacks(new SyncMcpToolCallbackProvider(mcpSyncClients))
        .build();

This setup allows the AI model to:

Understand when to use Brave Search
Format queries appropriately
Process and incorporate search results into responses

Dependencies

The project uses:

Spring Boot 3.3.6
Spring AI 1.0.0-SNAPSHOT
spring-ai-starter-model-openai
spring-ai-starter-mcp-client

3.3 KiB Raw Permalink Blame History