# Installation Guide
**Connect RivalSearchMCP to Your AI Assistant in Minutes**
This guide shows you exactly how to connect RivalSearchMCP to different AI assistants and applications. No technical knowledge required!
## What You're Actually Doing
When you connect RivalSearchMCP to your AI assistant, you're telling it "Hey, when I ask for current information, use this tool to search the web instead of saying you don't know."
## Option 1: Use the Live Server (Easiest - No Installation)
**This is the recommended way for most users.**
### For Cursor Users
Click this button to add RivalSearchMCP to Cursor in one click:
[](cursor://anysphere.cursor-deeplink/mcp/install?name=RivalSearchMCP&config=eyJ1cmwiOiJodHRwczovL1JpdmFsU2VhcmNoTUNQLmZhc3RtY3AuYXBwL21jcCJ9)
**What happens:**
1. Click the button
2. Cursor opens and asks if you want to add RivalSearchMCP
3. Click "Yes" or "Add"
4. Your AI assistant can now search the web!
### For Claude Desktop
1. Open Claude Desktop
2. Go to **Settings** → **MCP Servers**
3. Click **Add Server**
4. Enter this URL: `https://RivalSearchMCP.fastmcp.app/mcp`
5. Click **Save**
6. Restart Claude Desktop
### For VS Code
1. Open VS Code
2. Create or open `.vscode/mcp.json` in your workspace
3. Add this configuration:
```json
{
"mcpServers": {
"RivalSearchMCP": {
"url": "https://RivalSearchMCP.fastmcp.app/mcp"
}
}
}
```
### For Other MCP-Compatible Apps
Add this to your MCP configuration:
```json
{
"mcpServers": {
"RivalSearchMCP": {
"url": "https://RivalSearchMCP.fastmcp.app/mcp"
}
}
}
```
## Option 2: Install Locally (For Advanced Users)
**Only do this if you want full control or the live server doesn't work.**
### Step 1: Install Python
- **Windows**: Download from [python.org](https://python.org) - check "Add Python to PATH"
- **macOS**: Download from [python.org](https://python.org) or use `brew install python`
- **Linux**: `sudo apt install python3 python3-pip` (Ubuntu/Debian)
### Step 2: Download RivalSearchMCP
```bash
git clone https://github.com/rivalsearchmcp/rivalsearchmcp.git
cd rivalsearchmcp
```
### Step 3: Install Dependencies
```bash
pip install -r requirements.txt
```
### Step 4: Run the Server
```bash
python -m src.server
```
### Step 5: Connect Your AI Assistant
Use this configuration instead of the live server URL:
```json
{
"mcpServers": {
"rivalsearchmcp": {
"command": "python",
"args": ["-m", "src.server"],
"transport": {
"type": "stdio"
}
}
}
}
```
## Test Your Connection
### Simple Test
Ask your AI assistant: **"Search for 'latest AI news' and tell me what you find"**
**If it works:** Your AI will search the web and give you current information.
**If it doesn't work:** Your AI will say it can't access current information or doesn't know how to search.
### What You Should See
- **Before**: "I can't access current information" or "I don't know about recent events"
- **After**: Your AI searches the web and gives you current articles, news, and information
## Troubleshooting
### "Server not found" or "Connection failed"
- Make sure you're using the correct URL: `https://RivalSearchMCP.fastmcp.app/mcp`
- Check your internet connection
- Try restarting your AI assistant
### "Tool not available" or "Function not found"
- Make sure the server is connected properly
- Try asking: "What tools do you have available?"
- Restart your AI assistant
### Local installation issues
- Make sure Python is installed: `python --version`
- Make sure you're in the right directory
- Check that dependencies are installed: `pip list`
## What Your AI Can Do Now
Once connected, your AI assistant can:
- **Search the web** for current information
- **Read websites** and summarize content
- **Get latest news** and updates
- **Research topics** using multiple sources
- **Find current prices** and data
- **Analyze trends** and patterns
## Need Help?
- **Connection issues?** Try the live server first
- **Still not working?** [Report an issue](https://github.com/rivalsearchmcp/rivalsearchmcp/issues)
- **Want to learn more?** Check the [User Guide](../user-guide/overview.md)
---
**Ready to test?** Ask your AI assistant to search for something current!
