OpenDia
About
An open-source server that exposes browser functions via MCP, allowing AI models to interact with browser capabilities.
Details
- Author
- aaronjmars
- Categories
- Developer Tools, Automation, Other
Jump to
Setup
Install OpenDia in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/aaronjmars/opendia
Follow the installation instructions in the repository README, then restart your MCP client.
The open alternative to Dia / Perplexity Comet.
Connect your browser to AI models — no browser switching needed.
Works seamlessly with Chrome, Firefox, and any Chromium browser. Private, local-first & MCP focused.
New to MCPs, or not technical? We recommendPerplexity Comet.
- 🔄Universal AI Support: Works with Claude, ChatGPT, Cursor, Grok Build and even local models
- 🎯Anti-Detection: Specialized bypasses for Twitter/X, LinkedIn, Facebook
- 📱Smart Automation: AI understands your pages and finds the right elements
- 🛡️Privacy-First: Everything runs locally, your data stays with you
- ⚡Zero Setup: Get started with one command
Works withChrome, Firefox, and any Chromium-based browser:
- ✅Google Chrome
- ✅Arc
- ✅Mozilla Firefox
- ✅Microsoft Edge
- ✅Brave
- ✅Opera
- ✅Any Chromium based browser
Also perfect forCursor userswho want to automate their local testing and development workflows!
OpenDia exposes a standard MCP server, so it works with any client that speaks the protocol.
- Claude Desktop— paste the config below, restart.
- Claude Code— same MCP config, runs from any terminal.
- Cursor— add via Settings → MCP, or paste the same JSON.
- Grok Build— install the OpenDia plugin or add the MCP block to~/.grok/config.toml(seeGrok Buildbelow).
- ChatGPT— works with Auto-Tunnel mode (npx opendia --tunnel), then add the ngrok URL as a connector.
Should work(any MCP-compatible client): Windsurf, Zed, Continue, Cline, Goose, Open WebUI.
Wired it up with something else? Open a PR — the list grows as MCP grows.
- "Summarize all the articles I read today and post a Twitter thread about the key insights"
- "Find interesting articles related to AI from my bookmarks and create a reading list"
- "Read this article and post a thoughtful comment on the LinkedIn version"
- "Check my recent Twitter bookmarks and summarize the main themes"
- "Browse my latest emails and tell me what needs urgent attention"
- "Find all the GitHub repos I visited this week and create a summary report"
- "Extract the main points from this research paper and save them to my notes"
- "Search my browsing history for that article about AI safety I read last month"
🤖 Development & Testing (Perfect for Cursor!)
- "Test my web app's signup flow and take screenshots at each step"
- "Fill out this form with test data and check if validation works"
- "Navigate through my app and check if all the buttons work properly"
- "Use my connected wallet to test this DeFi interface"
- "Open tabs for all my daily news sources and summarize the top stories"
- "Draft replies to my unread messages based on the context"
- "Monitor this webpage and notify me when the content changes"
- "Automatically bookmark interesting articles I'm reading"
- "Apply a cyberpunk theme to this documentation site to make it more engaging"
- "Make this page dark mode with green text for late-night reading"
- "Add rainbow party effects to celebrate finishing this project"
- "Transform this boring form with a retro 80s theme while I fill it out"
- "Use high contrast styling so I can read this better"
- Downloadopendia-chrome-1.1.0.zipfromreleases
- Extract the zip file to a folder
- Go tochrome://extensions/(or your browser's extension page)
- Enable "Developer mode"
- Click "Load unpacked" and select the extracted folder
- Downloadopendia-firefox-1.1.0.zipfromreleases
- Extract the zip file to a folder
- Go toabout:debugging#/runtime/this-firefox
- Click "Load Temporary Add-on..."
- Select themanifest.jsonfile from the extracted folder
Note: Firefox extensions are loaded as temporary add-ons and will be removed when Firefox restarts. This is a Firefox limitation for unsigned extensions.
Option 1: Double-click Installation (Recommended)
- Download theopendia.dxtfile fromreleases
- Double-click the.dxtfile to install automatically
- The MCP will be added to your Claude Desktop configuration
Option 2: Manual ConfigurationAdd to your Claude Desktop configuration:
{ "mcpServers": { "opendia": { "command": "npx", "args": ["opendia"] } } }
For Cursor or other AI tools, use the same configuration or follow their specific setup instructions.
OpenDia ships a first-classGrok Build pluginunderplugins/opendia/(marketplace index:.grok-plugin/marketplace.json).
grok plugin marketplace add aeonfun/opendia grok plugin install opendia --trust
Enable the plugin in/plugins(trust is required for MCP). Details, skills, and troubleshooting:plugins/opendia/README.md.
Add to~/.grok/config.toml(or a project.grok/config.toml):
[mcp_servers.opendia] command = "npx" args = ["-y", "opendia"] enabled = true startup_timeout_sec = 60
Then install the browser extension and confirm it is connected tows://localhost:5555.
- Chrome extension: ws://localhost:5555 (auto-discovery enabled)
- Claude Desktop: stdio (existing config)
- Local SSE:http://localhost:5556/sse
# Use custom ports npx opendia --port=6000 # Uses 6000 (WebSocket) + 6001 (HTTP) npx opendia --ws-port=5555 --http-port=5556 # Specify individually # Handle port conflicts # Note: Existing OpenDia processes are automatically terminated on startup
- Automatically creates ngrok tunnel
- Copy URL for ChatGPT/online AI services
- Local functionality preserved
Because the tunnel is public,/sserequires a bearer token in this mode. The server prints one at startup; pass--token=<value>to pin a fixed one across restarts. Send it with every request:
curl -H "Authorization: Bearer <token>" https://<tunnel>.ngrok-free.app/sse
Both listeners bind127.0.0.1by default, andngrokconnects from this machine, so the tunnel works without widening anything. To expose the HTTP/SSE port on the LAN deliberately:
npx opendia --http-host=0.0.0.0 # requires a token, same as --tunnel
The WebSocket extension channel always stays on loopback.
Note: For auto-tunneling to work, you need ngrok installed:
# Using Chocolatey choco install ngrok # Or download from https://ngrok.com/download
# Ubuntu/Debian curl -s https://ngrok-agent.s3.amazonaws.com/ngrok.asc | sudo tee /etc/apt/trusted.gpg.d/ngrok.asc >/dev/null echo "deb https://ngrok-agent.s3.amazonaws.com buster main" | sudo tee /etc/apt/sources.list.d/ngrok.list sudo apt update && sudo apt install ngrok # Or download from https://ngrok.com/download
Then get your free authtoken fromhttps://dashboard.ngrok.com/get-started/your-authtokenand run:
ngrok config add-authtoken YOUR_TOKEN_HERE
OpenDia gives AI models18 powerful browser tools:
- Analyze any webpage- AI automatically finds buttons, forms, and interactive elements
- Extract content intelligently- Get clean text from articles, social posts, or search results
- Understand context- AI knows what type of page it's looking at and how to interact with it
- Click anything- Buttons, links, menus - AI finds and clicks the right elements
- Fill forms smartly- Works even on complex sites like Twitter, LinkedIn, Facebook
- Navigate seamlessly- Go to pages, scroll, wait for content to load
- Handle modern web apps- Bypasses detection on social platforms
- Multi-tab workflows- Open, close, switch between tabs automatically
- Organize your workspace- Let AI manage your browser tabs efficiently
- Coordinate complex tasks- Work across multiple sites simultaneously
- Bookmarks & History- Find that article you read last week
- Current page content- Get selected text, links, or full page content
- Real-time information- Work with whatever's currently on your screen
- Social media posting- Bypass automation detection on Twitter/X, LinkedIn, Facebook
- Natural interactions- Mimics human behavior to avoid triggering security measures
- Reliable automation- Works consistently even on sites that block typical automation tools
Once everything is set up, try asking your AI:
"Read the article on this page and create a Twitter thread summarizing the main points"
"Look through my browser history from this week and find articles about machine learning. Summarize the key trends."
"Check my Twitter bookmarks and organize them into categories. Create a summary of each category."
"Open tabs for my usual morning reading sites and give me a briefing of today's top stories"
"Fill out this contact form with test data and check if the submission works properly"
"Find that GitHub repo I was looking at yesterday about React components and bookmark it for later"
"Apply a dark hacker theme to this page to make it look more interesting"
"Make this boring documentation page feel like a cozy coffee shop"
"Add some matrix rain effects to this page for 30 seconds for a cool screenshot"
"Transform this page with a high contrast theme for better readability"
graph LR A[AI Model] --> B[OpenDia Server] B --> C[Browser Extension] C --> D[Your Browser] D --> E[Any Website]
- You askyour AI to do something browser-related
- AI callsOpenDia tools to understand and interact with pages
- OpenDia controlsyour browser through the extension
- You get results- AI can see what happened and respond intelligently
- ✅Everything runs locally- No cloud processing of your browsing data
- ✅You control access- Extension only works when you want it to
- ✅Open source- Full transparency of what the code does
- ✅No tracking- We don't collect or store any of your information
Important: This tool requires broad browser permissions to function. Only use with AI models you trust, and in environments where you're comfortable with browser automation.
Love to have your help making OpenDia better!
git clone https://github.com/aeonfun/opendia.git cd opendia # Start the server cd opendia-mcp npm install npm start # Load extension in your browser # Chrome: Go to chrome://extensions/ → Developer mode → Load unpacked: ./opendia-extension/dist/chrome # Firefox: Go to about:debugging#/runtime/this-firefox → Load Temporary Add-on → ./opendia-extension/dist/firefox/manifest.json # Extension will auto-connect to server on localhost:5555
- 🐛Report bugsviaGitHub Issues
- 💡Share it on social medias
- 🔧Add new browser capabilities
- 📖Improve documentation
- 🧪Test with different AI models
Ready to supercharge your browser with AI? Get started withnpx opendia! 🚀
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
Official Chrome DevTools MCP server for controlling and inspecting a live Chrome browser from coding agents such as Gemini, Claude, Cursor, and Copilot.
Automate your browser by connecting to Chrome's debugging port, preserving your login state.
AI-controlled browser farm — run N isolated Chromium instances with 36 MCP tools for device emulation, screenshots, network/geolocation simulation, and diagnostics. Cross-platform: Windows GUI + headed/headless on Linux/macOS.
A Chrome DevTools Extension that enables browser automation through the Model Context Protocol (MCP) for AI applications.
MCP server for browser automation via Puppeteer
Native desktop + browser automation MCP server with 82 tools — accessibility APIs (macOS/Windows), Chrome DevTools Protocol, anti-detection, memory, jobs, and reusable playbooks.
A faster, parallel browser for AI agents: it drives your own Chrome from Claude Desktop or any MCP client, on fewer round trips and tokens than Playwright MCP.
Browser automation via MCP for Chrome and Firefox
Bring the full power of BrowserStack’s Test Platform to your AI tools, making testing faster and easier for every developer and tester on your team.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
