mcpField
M
MCP Desktop DVR

MCP Desktop DVR

Local
github.com/Randroids-DojoProductivity4Updated 73w ago
Repository

About MCP Desktop DVR

MCP Desktop DVR captures desktop activity on macOS in a 30-minute rolling buffer and provides analysis of selected time segments. It is for users who want Claude to inspect screen activity, with cloud, local, and OCR analysis options.

Key features

  1. 01Maintain a 30-minute rolling buffer
  2. 02Capture a specific application window
  3. 03Extract and analyze time segments
  4. 04Analyze with GPT-4o, Tarsier2-7B, or OCR
  5. 05Start, stop, and configure capture

Use cases

  1. 01Review recent screen activity with Claude
  2. 02Identify errors visible on screen
  3. 03Analyze a workflow captured in a time segment

MCP support

Exposes MCP tools for starting and stopping capture, checking status, configuring capture, and analyzing recent desktop activity.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MCP Desktop DVR

✅ PRODUCTION READY

MCP server for desktop video capture and analysis on macOS. Enables Claude to analyze screen activity through a 30-minute rolling buffer.

Features

  • 30-minute rolling buffer of desktop activity
  • AI-powered visual analysis with OpenAI GPT-4o (cloud) or Tarsier2-7B (local)
  • Intelligent fallback from OpenAI → Tarsier → OCR
  • Window-specific capture by application
  • Precise extraction of time segments
  • Production-ready with comprehensive testing

Requirements

  • macOS 10.15+
  • Node.js 18+
  • Screen Recording permission

Quick Setup

npm install && npm run build

Add to Claude Desktop config:

{
  "mcpServers": {
    "desktop-dvr": {
      "command": "node",
      "args": ["/path/to/mcp-desktop-dvr/dist/index.js"],
      "env": {
        "ANALYZER_PREFERENCE": "auto"  // Auto-select best available analyzer
      }
    }
  }
}

Tools

  • start-continuous-capture - Begin recording
  • analyze-desktop-now - Extract and analyze recent activity
  • get-capture-status - Check recording status
  • stop-capture - Stop recording
  • configure-capture - Update settings

Usage

Ask Claude:

  • "Start desktop recording"
  • "Analyze the last 30 seconds"
  • "What errors are on my screen?"

Video Analysis Options

The analyze-desktop-now tool supports multiple analyzers with different strengths:

☁️ OpenAI GPT-4o Vision (Primary - Cloud)
  • Most accurate analysis using GPT-4o via Responses API
  • Automatic GIF segmentation - creates 10-second segments from videos
  • Smart file management - preserves all GIF files locally alongside MP4
  • Efficient uploads - sends only first segment to OpenAI for analysis
  • Comprehensive understanding of complex workflows
  • Requires OpenAI API key and internet connection
  • Fast processing (~3-5 seconds)
🤖 Tarsier2-7B (Fallback - Local AI)
  • Runs locally on your Mac using Metal Performance Shaders
  • No API keys required - completely private
  • Excellent visual understanding of UIs, games, and graphics
  • Moderate processing time (~5-15 seconds for 30-second clips)
📝 OCR Text Extraction (Final Fallback)
  • Basic text extraction from screenshots
  • Limited effectiveness with modern UIs
  • Always available as final fallback

Configuration

Set environment variables in your MCP config:

"env": {
  "ANALYZER_PREFERENCE": "auto",       // Options: "auto", "openai", "tarsier", "ocr"
  "OPENAI_API_KEY": "sk-...",          // For OpenAI (recommended)
  "OPENAI_MODEL": "gpt-4o"             // OpenAI model (default: gpt-4o)
}
Analyzer Selection Logic
  • "auto" (default) - Uses OpenAI if API key set, otherwise Tarsier, then OCR
  • "openai" - Forces OpenAI GPT-4o Vision (requires API key)
  • "tarsier" - Forces Tarsier2-7B local AI
  • "ocr" - Forces basic OCR text extraction

Documentation

License

ISC

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.