opentabs  by opentabs-dev

API-driven browser automation for AI agents

Created 2 months ago
319 stars

Top 85.1% on SourcePulse

GitHubView on GitHub
Project Summary

Browser automation clicks buttons. OpenTabs calls APIs. OpenTabs addresses the brittleness and inefficiency of traditional browser automation by directly interacting with web application APIs instead of simulating user interface actions. It targets developers and power users seeking reliable, fast, and programmatic control over web services, offering a significant performance and stability advantage over DOM scraping or visual automation tools.

How It Works

The core innovation lies in OpenTabs' ability to intercept and replay internal API calls made by a web application's frontend through the user's browser session. This bypasses the need for UI interaction, screenshots, or DOM parsing. By leveraging existing authenticated sessions, it achieves deterministic, high-speed execution, exposed as "MCP tools" that can be invoked by AI agents or scripts.

Quick Start & Requirements

  • Installation: Requires Node.js 22+ and Google Chrome. Install the CLI globally via npm install -g @opentabs-dev/cli.
  • Execution: Run opentabs start to initiate the MCP server. The first run generates configuration blocks for AI tools.
  • Browser Extension: Load the extension from ~/.opentabs/extension via chrome://extensions/ (enable Developer mode).
  • Plugin Management: Install plugins with opentabs plugin install <plugin-name>.
  • Setup Time: Estimated at five minutes.
  • Resources: Official documentation and quick-start guides are available at https://opentabs.dev/docs.

Highlighted Details

  • Extensive Ecosystem: Features over 100 plugins and approximately 2,000 tools, enabling direct interaction with services like Discord, Slack, GitHub, Jira, Notion, Figma, AWS, Stripe, Robinhood, and Panda Express.
  • Universal Browser Tools: Includes built-in capabilities for screenshots, clicking, typing, and network capture applicable to any browser tab without requiring specific plugins.
  • AI-Driven Plugin Development: Facilitates plugin creation by allowing users to point an AI at a website; the AI analyzes the page, discovers APIs, scaffolds code, and registers the plugin.
  • Robust Security Model: All plugins are disabled by default. AI-assisted code review is performed before enabling plugins. Permissions reset upon plugin updates, with three levels: Off, Ask (confirmation), or Auto, configurable per-plugin or per-tool. Runs entirely locally.

Maintenance & Community

The project is notable for being built entirely by AI agents, with all Product Requirement Documents (PRDs) open-sourced. Community support is available via Discord (https://discord.gg/opentabs).

Licensing & Compatibility

The project is licensed under the MIT license. It includes a disclaimer stating it is not affiliated with or endorsed by any third-party service. The MIT license generally permits commercial use and integration into closed-source projects.

Limitations & Caveats

The reliance on internal web application APIs means that OpenTabs' functionality is subject to changes in those APIs, potentially requiring plugin updates. No specific limitations regarding alpha/beta status or unsupported platforms were detailed in the provided README snippet.

Health Check
Last Commit

1 day ago

Responsiveness

Inactive

Pull Requests (30d)
45
Issues (30d)
3
Star History
316 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.