Vibium: Browser Automation for AI and Humans

🤖 Vibium: AI-Native Browser Automation

Created by the original author of Selenium, Vibium is a browser automation tool designed for both AI agents and developers. Its promise: give an AI agent a full browser through simple CLI commands or an MCP server.

⚡ Why it stands out:

• 🧠 AI-native — Installs as a “skill” that agents learn instantly
• 🪶 Zero config — A single ~10MB binary, Chrome downloads automatically
• 📐 Open standards — Built on WebDriver BiDi, no proprietary protocols
• 🔌 Multi-interface — CLI, MCP server, library for JS/TS, Python, and Java

🖥️ Basic workflow:

vibium go https://site.com  # navigate
vibium map                   # map elements → @e1, @e2
vibium click @e1             # interact by reference
vibium text                  # extract text

MCP integration with Claude Code, Gemini CLI, etc.:

claude mcp add vibium -- npx -y vibium mcp

Available on npm, PyPI, and Maven Central. Apache 2.0 license.

💡 Quick explanation

MCP (Model Context Protocol) is a standard for AI agents to access external tools. Vibium implements this protocol for web browsers, allowing an AI to browse the internet, fill forms, and extract information just like a human would.

More information at the link 👇

GitHub - VibiumDev/vibium: Browser automation for AI agents and humans

Browser automation for AI agents and humans. Contribute to VibiumDev/vibium development by creating an account on GitHub.

github.com ↗