Dataloupe

Overview

Dataloupe is an offline MCP server that lets you explore and analyze data files directly within your AI workflow. It handles CSV, JSON, Parquet, and Excel formats, offering preview, query, diff, and interactive HTML rendering capabilities. Any MCP-compatible AI assistant can invoke Dataloupe's tools to inspect and manipulate your data as part of a conversation or agent run.

The server runs locally via Docker using the image ghcr.io/aurelio-nakamura/dataloupe:0.13.0. No authentication or external credentials are needed — it starts immediately and communicates with your MCP client over stdio. Version 0.13.0 is available under the MIT license, with full source code at aurelio-nakamura/dataloupe on GitHub.

Dataloupe is categorized as File System on mcp.site and integrates with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and others. If you maintain this server, claim the listing to verify ownership and ensure the information stays accurate.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "dataloupe": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/aurelio-nakamura/dataloupe:0.13.0"
      ]
    }
  }
}