INSTALL_CLAUDE_DESKTOP.md•2.38 kB
# Event Horizon MCP Server - Claude Desktop Installation
## Quick Setup (Just Copy & Paste!)
### Step 1: Add to Claude Desktop Config
1. **Open Claude Desktop**
2. **Go to Settings** (gear icon in bottom left)
3. **Click "Developer"** tab
4. **Click "Edit Config"** button
5. **Paste this configuration:**
```json
{
"mcpServers": {
"event-horizon": {
"command": "node",
"args": [
"C:/Users/Manas Pandey/Documents/github/EventHorizon/dist/index.js"
]
}
}
}
```
### Step 2: Restart Claude Desktop
- Close Claude Desktop completely
- Reopen Claude Desktop
- You should see a small hammer/tools icon in the input area when the server is connected
## Verify Installation
1. Start a new conversation in Claude Desktop
2. Type: "What MCP tools do you have available?"
3. You should see 5 Event Horizon tools:
- `start_evolution`
- `add_solution`
- `score_solution`
- `evolve_generation`
- `get_evolution_status`
## Quick Test
Try this in Claude Desktop:
```
Use the start_evolution tool to begin evolving a solution for this problem:
Problem: "Create a simple recipe for chocolate chip cookies"
Consistency Checks:
- "Recipe must include exact measurements"
- "Instructions must be clear and easy to follow"
- "Recipe should work for beginners"
```
## Troubleshooting
**Issue**: Tools not showing up
- **Solution**: Make sure you completely restarted Claude Desktop after adding the config
**Issue**: "Cannot find module" error
- **Solution**: Verify the file path `C:/Users/Manas Pandey/Documents/github/EventHorizon/dist/index.js` exists
**Issue**: Permission errors
- **Solution**: The file already has correct permissions (executable), but if needed run:
```bash
chmod +x "C:/Users/Manas Pandey/Documents/github/EventHorizon/dist/index.js"
```
## What This Server Does
The Event Horizon MCP Server provides evolutionary solution generation tools that allow Claude to:
1. **Start Evolution**: Define a problem and evaluation criteria
2. **Generate Solutions**: Create multiple solution candidates
3. **Score Solutions**: Evaluate each solution against consistency checks
4. **Evolve Generations**: Use genetic algorithm principles to improve solutions
5. **Track Progress**: Monitor the evolution process and results
This creates a powerful system for iteratively improving solutions to complex problems!