Discover and explore top open-source AI tools and projects—updated daily.
sleepinginsummerBrowser control CLI for AI agents
New!
Top 81.8% on SourcePulse
Summary
agent-browser-cli provides a command-line interface and Chrome extension to expose real browser sessions to AI agents, addressing the need for precise web interaction while preserving login states and cookies. It benefits AI developers, automation testers, and data scrapers by enabling tasks like page scanning, JS execution, CDP control, and screenshots within agent workflows.
How It Works
A Rust CLI interfaces with a Chrome MV3 extension (tmwd_cdp_bridge) to control the user's active browser via the Chrome DevTools Protocol (CDP). This enables precise page manipulation, including JS execution, DOM reading, cookie management, and screenshots. Unlike Selenium/Playwright, it augments existing sessions, ideal for agents interacting with pre-authenticated or complex web applications.
Quick Start & Requirements
npm install -g @sleepinsummer/agent-browser-cli. Chrome extension: Manually load from release zip or source (assets/tmwd_cdp_bridge/).networkingMode=mirrored. Linux needs Chrome/Chromium supporting extension installation.AI_INSTALL.md for CLI, extension loading, and SKILL.md integration.Highlighted Details
alert/confirm/prompt dialogs during script execution.Maintenance & Community
Derived from the GenericAgent project. No specific community channels or maintainer details are provided in the README.
Licensing & Compatibility
Limitations & Caveats
Currently in beta (0.3.1-beta.1). WSL support requires specific configurations. The Chrome extension needs an active web tab for connection, and its UI auto-hides. Port management between CLI and extension requires careful configuration.
1 day ago
Inactive
h4ckf0r0day
vercel-labs