# TokenCut MCP Server

Use TokenCut from Claude Desktop, Cursor, or any MCP-compatible AI assistant.

## What is MCP?

The [Model Context Protocol (MCP)](https://modelcontextprotocol.io/) is an open standard that lets AI assistants use external tools. Our MCP server gives any compatible client the ability to compress text using TokenCut.

## Claude Desktop Setup

Add the following to your `claude_desktop_config.json`:

```
{
  "mcpServers": {
    "agentready": {
      "command": "npx",
      "args": ["-y", "agentready-mcp"],
      "env": {
        "AGENTREADY_API_KEY": "ak_your_api_key_here"
      }
    }
  }
}
```

## Cursor Setup

Same configuration — add to your Cursor MCP settings.

## Available Tools

### tokencut_compress

Compress text to reduce token count by 30-70%. Preserves code, URLs, and meaning.

| Parameter      | Type    | Default | Description                        |
|----------------|---------|---------|------------------------------------|
| text           | string  | required| Text to compress                    |
| level          | enum    | medium  | light, medium, or aggressive        |
| preserve_code  | boolean | true    | Keep code blocks intact             |
| preserve_urls  | boolean | true    | Keep URLs intact                    |
| target_model   | string  | gpt-4   | For cost estimation                 |

### tokencut_estimate

Quick estimation of compression potential without actually compressing.

## Pricing

- **Free tier:** 100,000 compressed tokens/month
- **Pay-as-you-go:** $0.08 per 1M compressed tokens

[Get Free API Key →](/content/signup/index.html)