Skip to content

Add MCP (Model Context Protocol) server support #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
webhead2oo9 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add MCP (Model Context Protocol) server support #4

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
182 changes: 182 additions & 0 deletions MCP_INTEGRATION.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,182 @@
# MCP (Model Context Protocol) Integration

This document describes how to configure and use MCP servers with the Discord AI Chat Bot.

## Overview

The Model Context Protocol (MCP) allows the bot to connect to external servers that provide:
- **Tools**: Additional functions that can be called by AI models
- **Resources**: Files and data that can be accessed during conversations
- **Prompts**: Reusable prompt templates

## Configuration

### 1. Define MCP Servers

Add MCP server configurations to your `config.json`:

```json
{
"mcpServerConfigurations": {
"filesystem": {
"transportType": "stdio",
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/allowed/directory"]
},
"github": {
"transportType": "stdio",
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-github"],
"env": {
"GITHUB_PERSONAL_ACCESS_TOKEN": "your-github-token"
}
},
"memory": {
"transportType": "stdio",
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-memory"]
},
"custom-server": {
"transportType": "sse",
"url": "http://localhost:3000/mcp"
}
}
}
```

### 2. Assign MCP Servers to Connectors

Add MCP servers to connector configurations:

```json
{
"connectorConfigurations": {
"ToolsOpenAIConnector": {
"class": "classes/connectors/ToolsOpenAIConnector",
"connectionOptions": {
"url": "https://api.openai.com/v1/chat/completions",
"apiKey": "OPENAI_KEY",
"tools": ["SearxingTool"],
"mcpServers": ["filesystem", "memory", "github"]
}
}
}
}
```

## Transport Types

### stdio Transport
- Used for local MCP servers that run as processes
- Requires `command` and optionally `args` and `env`
- Example servers: filesystem, GitHub, SQLite

### SSE (Server-Sent Events) Transport
- Used for remote MCP servers accessible via HTTP
- Requires `url` pointing to the SSE endpoint
- Suitable for cloud-hosted MCP servers

## Available MCP Servers

### Official Servers

1. **Filesystem** (`@modelcontextprotocol/server-filesystem`)
- Access files and directories
- Read/write operations
- Safe sandboxed access

2. **GitHub** (`@modelcontextprotocol/server-github`)
- Repository operations
- File reading/searching
- Issue and PR management

3. **Memory** (`@modelcontextprotocol/server-memory`)
- Key-value storage
- Persistent memory across conversations

4. **PostgreSQL** (`@modelcontextprotocol/server-postgres`)
- Database queries
- Schema inspection

5. **Brave Search** (`@modelcontextprotocol/server-brave-search`)
- Web search capabilities
- Requires Brave API key

### Community Servers

See [MCP Servers List](https://github.com/modelcontextprotocol/servers) for more options.

## How It Works

1. **Initialization**: When the bot starts, it connects to all configured MCP servers
2. **Tool Discovery**: The bot automatically discovers tools from each MCP server
3. **Tool Integration**: MCP tools are seamlessly integrated with existing bot tools
4. **Resource Access**: MCP resources can be accessed as attachments in responses
5. **Prompt Templates**: MCP prompts can enhance system instructions

## Features

### MCP Tools
- Automatically discovered and made available to AI models
- Work alongside existing bot tools (SearxingTool, WolframTool, etc.)
- Support the same interfaces as native tools

### MCP Resources
- Files and data from MCP servers
- Automatically converted to attachments in Discord messages
- Support for text, images, and other media types

### MCP Prompts
- Reusable prompt templates
- Can be used to enhance system instructions
- Support for parameterized prompts

## Example Use Cases

1. **File System Access**
```json
"mcpServers": ["filesystem"]
```
AI can read/write files in allowed directories

2. **GitHub Integration**
```json
"mcpServers": ["github"]
```
AI can browse repos, read code, manage issues

3. **Persistent Memory**
```json
"mcpServers": ["memory"]
```
AI can remember information across conversations

4. **Combined Powers**
```json
"mcpServers": ["filesystem", "github", "memory"]
```
Full suite of capabilities

## Security Considerations

- **Sandboxing**: Filesystem server only accesses allowed directories
- **Authentication**: Use environment variables for API keys
- **Transport Security**: SSE transport should use HTTPS in production
- **Tool Permissions**: Each connector only has access to its assigned MCP servers

## Troubleshooting

1. **Server won't start**: Check command path and arguments
2. **Tools not appearing**: Verify server is running and discoverable
3. **Authentication errors**: Ensure API keys are set in environment
4. **Connection timeouts**: Check firewall and network settings

## Development

To create a custom MCP server:
1. Implement the MCP protocol specification
2. Expose via stdio or SSE transport
3. Add configuration to `mcpServerConfigurations`
4. Assign to appropriate connectors

See [MCP Documentation](https://modelcontextprotocol.io) for protocol details.
7 changes: 7 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@ A powerful, multi-provider Discord bot for AI interactions. Seamlessly integrate
- Content moderation
- Configurable system instructions
- Passive listening with custom triggers
- MCP (Model Context Protocol) server integration

- **Discord Integration**
- Slash commands
Expand Down Expand Up @@ -53,6 +54,12 @@ A powerful, multi-provider Discord bot for AI interactions. Seamlessly integrate
- Add Reactions
- Use External Emojis

- **MCP Integration** 🔌
- Connect to MCP servers for extended capabilities
- Access external tools and resources
- Support for filesystem, GitHub, databases, and more
- See [MCP Integration Guide](./MCP_INTEGRATION.md) for details

## Quick Start 🚀

1. **Clone the Repository**
Expand Down
Loading