OpenDia provides an open-source solution for AI models to control web browsers, leveraging existing user sessions, accounts, and browser data. It targets users who want to automate browser tasks, from content creation and research to development testing, without context switching or starting from scratch. The primary benefit is seamless integration with a user's digital life for AI-driven browser automation.

How It Works

OpenDia operates via a local server that communicates with a browser extension. The AI model interacts with the OpenDia server, which then uses the browser extension to perform actions like clicking elements, filling forms, and extracting data. This architecture allows AI to access and utilize the user's logged-in sessions, cookies, and extensions, enabling sophisticated automation and anti-detection capabilities on sites like Twitter/X and LinkedIn.

Quick Start & Requirements

1. Install the browser extension (Chrome/Chromium, Firefox) by loading the unpacked extension files from releases.
2. Connect to your AI by downloading the .dxt file for Claude Desktop or manually configuring your AI tool's MCP settings to point to the OpenDia server (defaulting to npx opendia).
3. Run npx opendia locally. For external AI services, use npx opendia --tunnel (requires ngrok installation and authentication).

Highlighted Details

• Supports universal AI integration with models like Claude, ChatGPT, Cursor, and local LLMs.
• Offers specialized anti-detection features for social media platforms (Twitter/X, LinkedIn, Facebook).
• Provides 18 browser tools for smart page understanding, natural interactions, tab management, and data access.
• Enables visual customization of webpages with various themes and effects.

Maintenance & Community

The project is maintained by aaronjmars. Contributions are welcomed via GitHub Issues for bug reporting, adding capabilities, improving documentation, and testing.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source applications.

Limitations & Caveats

The project requires broad browser permissions, and users should only trust the AI models they connect with. Firefox extensions are loaded as temporary add-ons and may be removed upon browser restart unless a more permanent solution is implemented.