# datoon MCP server

> Converts structured data like JSON, CSV, and Parquet into the TOON format to reduce LLM token usage based on estimated savings.

- Repository: https://github.com/andrii-su/datoon
- Page: https://hostedmcp.org/servers/andrii-su/datoon
- Transport: stdio
- Stars: 2
- License: MIT
- Latest version: 3.0.0

## Install

### Claude Code

```sh
claude mcp add datoon -- uvx datoon
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "datoon": {
      "command": "uvx",
      "args": [
        "datoon"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "datoon": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "datoon"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"datoon","type":"stdio","command":"uvx","args":["datoon"]}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "datoon": {
      "command": "uvx",
      "args": [
        "datoon"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add datoon -- uvx datoon
```

### Codex config.toml (`~/.codex/config.toml`)

```toml
[mcp_servers.datoon]
command = "uvx"
args = ["datoon"]
```

### Gemini CLI

```sh
gemini mcp add -s user datoon uvx -- datoon
```

### Gemini settings.json (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "datoon": {
      "command": "uvx",
      "args": [
        "datoon"
      ]
    }
  }
}
```

## FAQ

### What formats does datoon support?

It supports JSON, CSV, JSONL, YAML, XML, Parquet, Avro, ORC, Excel, and Apple Numbers.

### How does the decision to convert work?

The server analyzes the data structure and only converts to TOON if it meets a savings threshold and structural requirements, otherwise it keeps the original format.

### Can I use different tokenizers for estimation?

Yes, you can use local counters like tiktoken or character counts, or network-based counters like Claude or Hugging Face tokenizers.
