Content
# Ollama-MCP Bridge WebUI
A TypeScript implementation that connects local LLMs (via Ollama) to Model Context Protocol (MCP) servers with a web interface. This bridge allows open-source models to use the same tools and capabilities as Claude, enabling powerful local AI assistants that run entirely on your own hardware.
## Features
- **Multi-MCP Integration**: Connect multiple MCP servers simultaneously
- **Tool Detection**: Automatically identifies which tool to use based on queries
- **Web Interface**: Clean UI with collapsible tool descriptions
- **Comprehensive Toolset**: Filesystem, web search, and reasoning capabilities
## Setup
### Automatic Installation
The easiest way to set up the bridge is to use the included installation script:
```bash
./install.bat
```
This script will:
1. Check for and install Node.js if needed
2. Check for and install Ollama if needed
3. Install all dependencies
4. Create the workspace directory (../workspace)
5. Set up initial configuration
6. Build the TypeScript project
7. Download the Qwen model for Ollama
After running the script, you only need to:
1. Add your API keys to the `.env` file (the `$VARIABLE_NAME` references in the config will be replaced with actual values)
### Manual Setup
If you prefer to set up manually:
1. Install Ollama from [ollama.com/download](https://ollama.com/download)
2. Pull the Qwen model: `ollama pull qwen2.5-coder:7b-instruct-q4_K_M`
3. Install dependencies: `npm install`
4. Create a workspace directory: `mkdir ../workspace`
5. Configure API keys in `.env`
6. Build the project: `npm run build`
## Configuration
The bridge is configured through two main files:
### 1. bridge_config.json
This file defines MCP servers, LLM settings, and system prompt. Environment variables are referenced with `$VARIABLE_NAME` syntax.
Example:
```json
{
"mcpServers": {
"filesystem": {
"command": "node",
"args": [
"To/Your/Directory/Ollama-MCP-Bridge-WebUI/node_modules/@modelcontextprotocol/server-filesystem/dist/index.js",
"To/Your/Directory/Ollama-MCP-Bridge-WebUI/../workspace"
],
"allowedDirectory": "To/Your/Directory/Ollama-MCP-Bridge-WebUI/../workspace"
},
"brave-search": {
"command": "node",
"args": [
"To/Your/Directory/Ollama-MCP-Bridge-WebUI/node_modules/@modelcontextprotocol/server-brave-search/dist/index.js"
],
"env": {
"BRAVE_API_KEY": "$BRAVE_API_KEY"
}
},
"sequential-thinking": {
"command": "node",
"args": [
"To/Your/Directory/Ollama-MCP-Bridge-WebUI/node_modules/@modelcontextprotocol/server-sequential-thinking/dist/index.js"
]
}
},
"llm": {
"model": "qwen2.5-coder:7b-instruct-q4_K_M",
"baseUrl": "http://localhost:11434",
"apiKey": "ollama",
"temperature": 0.7,
"maxTokens": 8000
},
"systemPrompt": "You are a helpful assistant that can use various tools to help answer questions. You have access to three main tool groups: 1) Filesystem operations - for working with files and directories, 2) Brave search - for finding information on the web, 3) Sequential thinking for complex problem-solving. When a user asks a question that requires external information, real-time data, or file manipulation, you should use a tool rather than guessing or using only your pre-trained knowledge."
}
```
### 2. .env file
This file stores sensitive information like API keys:
```
# Brave Search API key
BRAVE_API_KEY=your_brave_key_here
```
The bridge will automatically replace `$BRAVE_API_KEY` in the configuration with the actual value from your `.env` file.
## Usage
### Starting the Bridge
Simply run:
```bash
./start.bat
```
This will start the bridge with the web interface.
### Web Interface
Open http://localhost:8080 (or the port shown in the console) in your browser to access the web interface.
## License
MIT License - See LICENSE file for details
MCP Store