MCP Screen Text
Allows the capture of application-specific screenshots from Safari and extraction of text through OCR capabilities.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@MCP Screen Text Extract the text from my active Chrome window"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP Screen Text
A Model Context Protocol (MCP) server that provides screen capture and optical character recognition (OCR) capabilities.
🎥 Demo Video
See MCP Screen Text in action - capturing screens and extracting text with Claude Desktop
Related MCP server: screenshot-mcp
Features
Screen Capture: Take screenshots of specific displays or applications
Application-Specific Screenshots: Capture screenshots of specific application windows
OCR Text Extraction: Extract text from screenshots or existing images
Desktop Storage: All screenshots are saved to a "Screenshots" folder on your Desktop
Multi-format Support: Support for PNG and JPG image formats
Multi-language OCR: Support for multiple languages in text recognition
Application Discovery: List running applications available for capture
Tools Available
capture_screen
Captures a screenshot of the entire screen or a specific display.
Parameters:
display(number, optional): Display number to capture (0 for primary display)format(string, optional): Image format for the screenshot ('png' or 'jpg')
capture_application_screen
Captures a screenshot of a specific application window.
Parameters:
applicationName(string, required): Name of the application to capture (e.g., 'Safari', 'Chrome', 'Finder')format(string, optional): Image format ('png' or 'jpg')
list_applications
Lists all running applications that can be captured.
Parameters: None
extract_text
Extracts text from an existing image file using OCR.
Parameters:
imagePath(string, required): Path to the image filelanguage(string, optional): Language for OCR recognition (e.g., "eng", "spa", "fra")
capture_screen_and_extract_text
Captures a screenshot and extracts text from it in one operation. This is a convenience tool that combines screen capture and OCR and can work with both full screen and application-specific capture.
Parameters:
display(number, optional): Display number to capture (0 for primary display) - ignored if applicationName is providedlanguage(string, optional): Language for OCR recognition (e.g., "eng", "spa", "fra")applicationName(string, optional): Name of the application to capture (e.g., 'Safari', 'Chrome'). If provided, captures only this application's window instead of full screen.
Installation
npm installDevelopment
# Build the project
npm run build
# Run in development mode
npm run dev
# Run the built version
npm startDependencies
@modelcontextprotocol/sdk: MCP SDK for server implementationscreenshot-desktop: Cross-platform screenshot capturesharp: High-performance image processingtesseract.js: OCR text extraction
Usage with MCP Client
This server can be used with any MCP-compatible client. Configure your client to connect to this server using stdio transport.
Example configuration for Claude Desktop:
{
"mcpServers": {
"screen-text": {
"command": "node",
"args": ["path/to/mcp-screen-text/dist/index.js"]
}
}
}License
ISC
Available Tools
5 toolscapture_application_screenC
Captures a screenshot of a specific application window
Inputs: applicationName (required), format.
capture_screenC
Captures a screenshot of the specified display
Inputs: display, format.
capture_screen_and_extract_textA
Captures a screenshot and extracts text from it in one operation. Can capture full screen or a specific application window.
Inputs: display, language, applicationName.
extract_textC
Extracts text from an image using OCR
Inputs: imagePath (required), language.
list_applicationsB
Lists all running applications that can be captured
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Multiple tools have overlapping purposes that could cause confusion. For example, capture_application_screen and capture_screen_and_extract_text both handle application window screenshots, while capture_screen and capture_screen_and_extract_text both handle full-screen captures. The descriptions help somewhat, but the boundaries between tools are unclear, especially for agents trying to choose between similar options.
All tool names follow a consistent snake_case pattern with clear verb_noun structures (e.g., capture_screen, extract_text, list_applications). There are no deviations in naming style, making the set predictable and easy to parse programmatically.
With 5 tools, the count is well-scoped for a screen text capture domain, though it feels slightly thin. Each tool appears to serve a purpose, but the overlap suggests some tools could be consolidated, making the count borderline but still reasonable for the server's functionality.
The tool set covers core operations like capturing screens, extracting text, and listing applications, but there are notable gaps. For instance, there's no way to manage or delete captured images, and the OCR functionality is split across tools without clear update or configuration options, which could lead to agent workarounds for full lifecycle handling.
Maintenance
Related MCP Connectors
OCR text from images, screenshots and scanned PDFs in 41 languages, with line positions.
Extract text from images, photos and screenshots with OCR in 100+ languages, with line positions.
Full-page, first-screen, mobile or element screenshots of any URL as PNG, JPEG or PDF.
Inkscry: OCR for screenshots, scans and photos in 100+ languages; text, confidence, line boxes.
Related MCP Servers
- AlicenseNot gradedqualityNot gradedmaintenanceEnables LLMs to capture and analyze screenshots of your screen, windows, or regions with smart detection capabilities. Features natural language queries, automatic window targeting, and text enhancement for UI debugging and visual inspection.2MIT
- AlicenseNot gradedqualityCmaintenanceA cross-platform MCP server that allows AI agents to capture screenshots of specific windows, displays, or regions for native application testing. It provides tools to list active windows and monitors, enabling precise visual verification and interaction during automated workflows.21 npm2MIT
- AlicenseNot gradedqualityDmaintenanceEnables screenshot capture and visual analysis using cloud or local vision models, with tools to describe screens, list windows, and analyze images.488 npm13MIT
- AlicenseNot gradedqualityDmaintenanceProvides screenshot capture and vision analysis tools that enable AI to see and analyze screen content on Windows, forming an automated capture-analyze pipeline.1MIT