MCP Screenshot

An MCP server that captures screenshots and performs OCR text recognition for Japanese and English, producing outputs in multiple formats.

typescript

22

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "kazuph-mcp-screenshot": {
      "command": "npx",
      "args": [
        "-y",
        "@kazuph/mcp-screenshot"
      ],
      "env": {
        "OCR_API_URL": "http://localhost:8000"
      }
    }
  }
}

You run a specialized MCP server that captures screenshots and runs OCR to extract text in Japanese and English. It can output results in multiple formats and is designed to be invoked from an MCP client to automate screen capture and text recognition tasks.

How to use

To use this server from an MCP client, you start the server component and issue capture requests through the client. You can request a screenshot of the left half, right half, or the full screen, and you can choose the output format you want (JSON, Markdown, vertical, or horizontal). The OCR engine will handle recognition, prioritizing yomitoku when available and falling back to Tesseract.js if needed. When you issue a capture, you’ll receive the recognized text in the selected format.

How to install

Prerequisites: you need Node.js and npm installed on your system.

Install the MCP server package using npm via npx.

npx -y @kazuph/mcp-screenshot

Additional notes

The server relies on an OCR API endpoint for yomitoku. You supply the API base URL through an environment variable named OCR_API_URL. By default, this is set to http://localhost:8000.

Available tools

capture

Takes a screenshot and performs OCR. Options include region to choose left, right, or full, and format to output json, markdown, vertical, or horizontal.