Auto-Snap MCP
Provides containerized deployment option for the Auto-Snap MCP service, allowing users to run the screenshot and document processing functionality in an isolated environment.
Enables native support for capturing screenshots and processing documents on Linux systems using X11 display server.
Offers experimental support for capturing screenshots and processing documents on macOS systems when using XQuartz.
Utilizes Python for server functionality, allowing users to check dependencies and run the Auto-Snap MCP server for screenshot and document processing.
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., "@Auto-Snap MCP Capture the current window and convert it to a PDF"
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.
Auto-Snap MCP 📸
Automated screenshot capture and document processing for MCP Clients
Turn your screenshots into PDFs automatically! Auto-Snap lets your MCP client capture windows, process documents, and create PDFs with simple natural language commands.
🚀 Quick Start
Which Option Should I Choose?
🐳 Docker → Zero setup, works everywhere (recommended)
🪟 WSL2 → Windows users with WSL2 installed
🐧 Linux → Native Linux
Setup Steps
1. Get Auto-Snap:
git clone https://github.com/your-repo/auto-snap-mcp
cd auto-snap-mcp2. Install dependencies (Linux only):
# Skip this step if using Docker
sudo apt install -y wmctrl xdotool tesseract-ocr
uv sync3. Configure Claude Desktop:
Edit ~/.claude/claude_desktop_config.json and add one of these:
🐳 Docker (Recommended - No setup needed!)
{
"mcpServers": {
"auto-snap-mcp": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"-e", "DISPLAY=:0",
"-v", "/tmp/.X11-unix:/tmp/.X11-unix:rw",
"-v", "${HOME}/auto-snap-captures:/app/captures:rw",
"mcp/auto-snap-mcp:latest"
]
}
}
}🪟 WSL2 (Windows with WSL2)
{
"mcpServers": {
"auto-snap-mcp": {
"command": "wsl.exe",
"args": [
"bash", "-c",
"cd /home/YOUR_WSL_USERNAME/auto-snap-mcp && DISPLAY=:0 /home/YOUR_WSL_USERNAME/.local/bin/uv run python server.py"
]
}
}
}🐧 Linux Native
{
"mcpServers": {
"auto-snap-mcp": {
"command": "uv",
"args": ["run", "python", "server.py"],
"cwd": "/home/YOUR_USERNAME/auto-snap-mcp",
"env": {"DISPLAY": ":0"}
}
}
}4. Restart Claude Desktop and verify it works:
"List all my open windows" ← Start with this to test the connection
"Capture this PDF and convert to images"
"Take 5 screenshots and make them into a PDF"
✅ Working? You should see your windows listed. If not, check the 🚨 Not Working? section below.
Related MCP server: Textin MCP Server
📁 Where Your Files Are Saved
By default, all your captures go to:
~/auto-snap-captures/Full path examples:
Linux:
/home/username/auto-snap-captures/WSL2:
/home/username/auto-snap-captures/Windows (Docker):
C:\Users\username\auto-snap-captures\
✨ What It Does & Commands to Try
📋 Document Capture
Screenshot any window or the entire screen
Capture multi-page documents automatically
Works with PDFs, presentations, web pages
🔍 Image Processing
Extract text from screenshots (OCR)
Enhance image quality automatically
Process multiple images at once
📄 PDF Creation
Convert screenshots to PDF instantly
Organize files with smart naming
Compress PDFs for smaller size
🎯 Commands to Try
"Capture this document as PDF" → Takes screenshots and creates a PDF
"Extract text from these images" → Runs OCR on screenshots
"Archive this presentation" → Screenshots all slides into one PDF
"List all my open windows" → See what you can capture
"Take 5 screenshots and make them into a PDF" → Multi-page capture
💡 Pro Tips:
Be specific: "Capture the Chrome window" works better than "take a screenshot"
Multi-step: "Screenshot this presentation and extract all the text"
Batch work: "Process all images in my Downloads and make PDFs"
Want to change where files are saved? See the 🎨 Customize Your Captures section below.
🛠️ System Support
✅ Linux (native X11)
✅ WSL2 (Windows apps from Linux)
⚠️ macOS (with XQuartz - experimental)
🚨 Not Working?
Common fixes:
# Check dependencies
uv run python -c "from capture import check_dependencies; print(check_dependencies())"
# Fix X11 display
export DISPLAY=:0
# Test the server
uv run python server.pyStill stuck?
Make sure the config path is correct:
~/.claude/claude_desktop_config.jsonRestart Claude Desktop after config changes
Check Claude Desktop logs for errors
🎨 Customize Your Captures
Set where files go:
export AUTO_SNAP_OUTPUT_DIR="$HOME/Documents/Screenshots"Organize by date:
export AUTO_SNAP_USE_DATE_SUBDIRS=true
export AUTO_SNAP_INCLUDE_TIMESTAMP=trueCustom file names:
export AUTO_SNAP_FILE_NAME_TEMPLATE="doc_{page:04d}"🌍 Multiple Languages
Add more OCR languages:
sudo apt install tesseract-ocr-spa # Spanish
sudo apt install tesseract-ocr-fra # French
sudo apt install tesseract-ocr-deu # GermanThen tell Claude: "Process this document in Spanish"
🔐 Privacy
Everything runs locally on your machine
No cloud services or uploads
Only captures what you ask for
Automatic cleanup of temp files
Ready to automate your screenshots? Install Auto-Snap and start talking to your MCP Clients about your documents! 🚀
Available Tools
10 toolscapture_document_pagesA
Capture multiple pages from a document window with automatic navigation.
Args:
window_id: Window ID containing the document
page_count: Number of pages to capture
output_dir: Directory to save captured pages
navigation_key: Key to press for navigation (Page_Down, Right, space)
delay_seconds: Delay between navigation and capture
Returns:
JSON string with capture results and list of captured files.
Inputs: window_id (required), page_count (required), output_dir, navigation_key, +1 more. Returns: result
capture_full_screenB
Capture screenshot of the entire screen.
Args:
output_path: Optional path to save the screenshot
Returns:
JSON string with capture results and file path.
Inputs: output_path. Returns: result
capture_windowA
Capture screenshot of a specific window.
Args:
window_id: Window ID to capture (from list_windows)
output_path: Optional path to save the screenshot
Returns:
JSON string with capture results and file path.
Inputs: window_id (required), output_path. Returns: result
check_system_dependenciesB
Check if all required system dependencies are installed.
Returns:
JSON string with dependency check results.
Inputs: no named parameters. Returns: result
convert_to_pdfA
Convert a list of images to a PDF document.
Args:
image_paths: List of image file paths to convert
output_path: Path for the output PDF file
title: Optional title for the PDF document
sort_files: Whether to sort files by name before conversion
Returns:
JSON string with conversion results.
Inputs: image_paths (required), output_path (required), title, sort_files. Returns: result
debug_window_detectionB
Comprehensive debugging information for window detection issues.
Returns:
JSON string with detailed diagnostics about PowerShell environment,
process enumeration, and window detection capabilities.
Inputs: no named parameters. Returns: result
directory_to_pdfB
Convert all images in a directory to a PDF document.
Args:
image_dir: Directory containing images
output_path: Path for the output PDF file
title: Optional title for the PDF document
pattern: File pattern to match (default: all files)
Returns:
JSON string with conversion results.
Inputs: image_dir (required), output_path (required), title, pattern. Returns: result
full_document_workflowB
Complete workflow: capture document pages, optionally process them, and convert to PDF.
Args:
window_id: Window ID containing the document
page_count: Number of pages to capture
output_pdf: Path for the final PDF file
capture_dir: Temporary directory for captures
title: Optional PDF title
navigation_key: Key for page navigation
delay_seconds: Delay between navigation and capture
process_images_flag: Whether to enhance images before PDF conversion
Returns:
JSON string with complete workflow results.
Inputs: window_id (required), page_count (required), output_pdf (required), capture_dir, +4 more. Returns: result
list_windowsA
List all available windows for screenshot capture.
Returns:
JSON string containing list of windows with their IDs, titles, and properties.
Inputs: no named parameters. Returns: result
process_imagesC
Process images in a directory with various operations.
Args:
image_dir: Directory containing images to process
operations: List of operations (enhance, ocr, resize, deduplicate)
ocr_language: Language for OCR processing (default: eng)
Returns:
JSON string with processing results.
Inputs: image_dir (required), operations, ocr_language. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Most tools have distinct purposes, but there is some overlap between capture_document_pages and full_document_workflow, as the latter encompasses the former's functionality. Similarly, convert_to_pdf and directory_to_pdf both handle PDF conversion, though with different input sources. The descriptions help clarify these distinctions, but an agent might occasionally misselect between these pairs.
All tool names follow a consistent verb_noun pattern using snake_case, such as capture_window, list_windows, and process_images. This uniformity makes the tool set predictable and easy to navigate, with no deviations in naming conventions across the server.
With 10 tools, the server is well-scoped for document capture and processing workflows. Each tool serves a clear purpose, from window listing and capture to image processing and PDF conversion, without unnecessary redundancy or gaps, making the count appropriate for the domain.
The tool set covers core workflows for document capture, processing, and PDF conversion comprehensively, including dependencies and debugging. However, there is a minor gap in direct OCR or text extraction tools, as process_images includes OCR but lacks standalone options, which might require workarounds for specific tasks.
Related MCP Connectors
Document conversion MCP server: PDF to Markdown, image OCR, spreadsheet parsing.
Documents to Markdown/PDF, OCR, video and audio transcripts, TTS, screenshots, domains, frames.
Screenshots, PDFs, image editing, QR codes, email validation and more through one hosted MCP server.
Screenshot, PDF and Open Graph image API for AI agents: capture pages, render PDFs, make OG images.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceA server that provides document processing capabilities using the Model Context Protocol, allowing conversion of documents to markdown, extraction of tables, and processing of document images.19MIT

Textin MCP Serverofficial
AlicenseCqualityDmaintenanceA server that enables OCR capabilities to recognize text from images, PDFs, and Word documents, convert them to Markdown, and extract key information.329MIT- AlicenseNot gradedqualityNot gradedmaintenanceA server that enables AI assistants to control a browser through tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.Apache 2.0
- AlicenseCqualityDmaintenanceA server that allows AI agents and clients to programmatically capture screenshots of websites with options for image format and viewport selection.121 npm9MIT