MCP Server Devops Bridge

🚀 MCP Server DevOps Bridger and Proxy System

This repository contains the code for the MCP (Mission Control Panel) Server DevOps Bridger, which is equipped with a powerful proxy system that allows AI models to create, manage, and coordinate long-running agents.

🚀 Quick Start

Please check the code examples in the examples folder to learn how to use this library. You can also refer to the official documentation for more detailed information.

✨ Features

The system enables AI models to:

1. Create new long-running agents with custom system prompts and tasks.
2. Send commands to existing agents.
3. Enable communication between agents through a messaging system.
4. Monitor and manage the lifecycle of agents.

🔧 Technical Details

Key Components

Agent Management

The system provides the following features:

• Create Agent: Create a new agent with custom system prompts and tasks.
• List Agents: View all running agents and their statuses.
• Send Command: Send instructions or queries to a specific agent.
• Subscribe to Topic: Allow an agent to listen for messages on a specific channel.
• Terminate Agent: Terminate an agent when it is no longer needed.

Inter-Agent Communication

Agents can communicate with each other in the following ways:

• Message Bus: A centralized messaging system that routes messages to subscribed agents.
• Topics: Channels that agents can publish to and subscribe to.
• Direct Commands: Send instructions to a specific agent.

System Tools

The system provides the following tools:

• Command Execution: Run system commands with appropriate security controls.
• Messaging: Send and receive messages between agents.
• Agent Management: Create, monitor, and terminate agents.

💻 Usage Examples

Creating and Coordinating Agents

// Get all agent-related tools
tools := ai.GetAllToolsAsOpenAI()

// Create an OpenAI client
client := openai.NewClient()

// Coordinate agents using OpenAI
messages := []openai.ChatCompletionMessageParamUnion{
    openai.SystemMessage(`You are the coordinator of the agents.`),
    openai.UserMessage("Create two agents and have them work together."),
}

// Call OpenAI and use our tools
params := openai.ChatCompletionNewParams{
    Model:    openai.F(openai.ChatModelGPT4o),
    Messages: openai.F(messages),
    Tools:    openai.F(tools),
}

// Process the response and handle tool calls
// ...

Example Agent Workflow

1. Create Agents: Create agents dedicated to different tasks.

   Tool: agent
   Parameters: {"id": "researcher", "system_prompt": "You are a research agent...", "task": "Find information about climate change."}
   

2. Subscribe to Topics: Have agents listen for relevant messages.

   Tool: subscribe_agent
   Parameters: {"agent_id": "writer", "topic": "research_results"}
   

3. Send Messages: Share information between agents.

   Tool: send_agent_message
   Parameters: {"topic": "research_results", "content": "Here is the information I found in my research..."}
   

4. Send Commands: Issue direct instructions to agents.

   Tool: send_command
   Parameters: {"agent_id": "writer", "command": "Summarize your research results in three paragraphs."}
   

Environment Variable Configuration

The following is an example of the project's environment variable configuration:

# MCP Server configuration
export MCP_SERVER_PORT=8080
export MCP_SERVER_HOST=localhost

# Database connection configuration
export DB_HOST=localhost
export DB_PORT=3306
export DB_NAME=mcp_db
export DB_USER=root
export DB_PASSWORD=secret

# Logging configuration
export LOG_LEVEL=info
export LOG_FILE=log.txt

Quick Launch

You can quickly launch the MCP server by following these steps:

1. Clone the repository:

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

2. Install dependencies:

   go mod tidy
   

3. Start the server:

   go run main.go
   

Documentation

• MCP Server API Documentation: API Documentation Link
• Agent System Documentation: Agent Documentation Link

Contribution Guidelines

If you wish to contribute to the project, please refer to the Contribution Guidelines.

📄 License

This software is protected by the MIT License.