Paragonos Misc Tools

Part of ParagonOS by Metarune Labs Pvt Ltd

License: Apache 2.0

A collection of miscellaneous tools for Claude, Cursor, and MCP-compatible clients. Includes OCR (Optical Character Recognition) capabilities leveraging the Google Gemini API. Extract text from images using file paths or base64 encoded strings through the Model Context Protocol.

Overview

This MCP server provides enterprise-grade OCR capabilities, enabling AI assistants to extract text from images with high accuracy using Google's advanced Gemini models.

Example Use Case:

Extract text from the following image:

Convert it to plain text, e.g., fbVk

Key Features

• File-based OCR: Extract text directly from image files on your local system
• Base64 OCR: Extract text from base64 encoded image strings
• High Accuracy: Utilizes Google's advanced Gemini models for superior text recognition
• Easy Integration: Simple MCP tools for seamless AI assistant integration
• Multi-format Support: Works with various image formats (PNG, JPG, JPEG, etc.)

Requirements

• Python 3.8+
• Google Gemini API Key
• FastMCP
• Claude Desktop or Cursor (or any MCP client)

Installation

1. Clone Repository

git clone https://github.com/Paragon-OS/paragonos-misc-tools-mcp.git
cd paragonos-misc-tools-mcp

2. Install UV (if needed)

macOS/Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

3. Install Dependencies

uv sync

4. Install Playwright Browser (for Notion tools)

After installing dependencies, you need to install the Chromium browser for Playwright:

playwright install chromium

MCP Configuration

Edit your Claude config (~/Library/Application Support/Claude/claude_desktop_config.json) or Cursor config (~/.cursor/mcp.json):

{
  "mcpServers": {
    "paragonos-misc-tools": {
      "command": "uv",
      "args": [
        "--directory",
        "/full/path/to/paragonos-misc-tools-mcp",
        "run",
        "misc-tools-mcp"
      ],
      "env": {
        "GEMINI_MODEL": "gemini-2.5-flash-preview-05-20",
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY"
      }
    }
  }
}

Windows:

{
  "mcpServers": {
    "paragonos-misc-tools": {
      "command": "uv",
      "args": [
        "--directory",
        "C:\\full\\path\\to\\paragonos-misc-tools-mcp",
        "run",
        "misc-tools-mcp"
      ],
      "env": {
        "GEMINI_MODEL": "gemini-2.5-flash-preview-05-20",
        "GEMINI_API_KEY": "YOUR_GEMINI_API_KEY"
      }
    }
  }
}

Available Tools

ocr_image_file

Performs OCR on a local image file.

• Parameter: image_file (string) - Absolute or relative path to the image file
• Returns: Extracted text from the image

ocr_image_base64

Performs OCR on a base64 encoded image.

• Parameter: base64_image (string) - Base64 encoded string of the image
• Returns: Extracted text from the image

read_notion_markdown

Extracts markdown content from a Notion published site URL.

• Parameter: notion_url (string) - The full URL of the Notion published site (must be notion.site, not notion.so)
• Returns: The extracted markdown content as a string

Usage Examples

• "Extract text from image at /path/to/screenshot.png"
• "Read the text from this image: [base64 string]"
• "What does the text say in captcha.jpg?"
• "OCR this receipt image and extract the total amount"
• "Read the markdown from this Notion page: https://example.notion.site/..."

About ParagonOS

ParagonOS is a comprehensive AI integration infrastructure platform developed by Metarune Labs Pvt Ltd, providing secure, reliable, and extensible bridges between AI assistants and modern services.

License

Licensed under the Apache 2.0 License. See LICENSE file for details.

---

Developed by Metarune Labs Pvt Ltd Part of the ParagonOS ecosystem