VibiumDev/vibium

The verification layer for coding agents

What it solves

Vibium provides a verification layer for AI coding agents, allowing them to interact with and verify their work in a web browser. It eliminates the need for complex manual browser setup and proprietary protocols, giving agents the ability to navigate pages, fill forms, and capture screenshots to ensure their code changes are functioning correctly.

How it works

Vibium operates as a lightweight binary that acts as a bridge between an LLM/Agent and a web browser. It can be integrated as a CLI skill, an MCP (Model Context Protocol) server, or through client libraries for JavaScript/TypeScript, Python, and Java. It uses the WebDriver BiDi protocol to communicate with the browser, providing semantic search (finding elements by text, label, or role) rather than relying solely on CSS selectors.

Who it’s for

It is designed for developers building AI agents (such as Claude Code or Gemini) that need to perform browser automation and verification tasks autonomously.

Highlights

  • AI-native integration: Can be installed as a skill for agents to instantly learn browser automation.
  • Zero configuration: Automatically downloads the required browser (Chrome) upon installation.
  • Standards-based: Built on WebDriver BiDi instead of proprietary protocols.
  • Multi-interface: Supports CLI, MCP server, and language-specific APIs (JS/TS, Python, Java).
  • Semantic element discovery: Find elements by visible text, ARIA roles, or form labels.
  • Lightweight: Distributed as a single ~10MB binary with no runtime dependencies.

Related

  • Project
    reticlehq/reticle

    A verification layer for AI coding agents that proofreads code by monitoring a running app's internal state and network calls to catch silent failures.

  • Project
    TestSprite/testsprite-cli

    A CLI tool that enables AI coding agents to verify their work by running tests against live applications in the cloud and fixing bugs using detailed failure bundles.

  • Project
    browser-use/browser-use

    A library and framework that enables AI agents to control a web browser like a human to automate tasks such as filling forms and extracting data.

  • Project
    browserwing/browserwing

    A browser automation platform that turns websites into structured data using a visual recorder, a CLI for AI agents, and native support for MCP and Skills protocols.

  • Project
    callstack/agent-device

    A tool that gives AI coding agents a live feedback loop to inspect, control, and verify mobile, TV, and desktop apps across iOS, Android, and other platforms.