This is an early release preview. You may encounter bugs.
Vibetest Use logo
Unclaimed

Tool interface

Vibetest Use

MCP server that spins up multiple Browser Use agents to QA-test your site for UI bugs and broken links.

D 22/100 GitHub score ? This grade is derived from GitHub signals, not user votes. Open for the full breakdown.
Discussion

01 / Summary

About Vibetest Use.

Vibetest Use is a Model Context Protocol (MCP) server that launches several Browser Use agents in parallel to test a website for UI bugs, broken links, accessibility problems, and other technical issues. It works against live sites and local development servers, so you can check a site as you build it from inside Claude Code or Cursor.

You register the vibetest-mcp binary as an MCP server with a Google API key, then prompt your coding assistant with a target URL, an agent count, and whether to run headless. The default is three agents in non-headless mode, and more agents cover more of the site. Agents run on gemini-2.0-flash via Playwright-installed Chromium.

  • Parallel agents: several Browser Use agents explore the site at once, with the count set in the prompt
  • Issue types: UI bugs, broken links, accessibility issues, and other technical problems
  • Target flexibility: any public URL or a local development server
  • Headless option: run with visible browsers or headless
  • MCP clients: Claude Code or Cursor, registered through each client's MCP settings

02 / Discussion CREDIBILITY-GATED

Discussion

Reading is open to everyone. Posting and voting need a verified identity or a GitHub grade of B or higher.

  • No discussions yet.

04 / Build

Build with Vibetest Use.

Browse the catalog for frameworks, tools, and harnesses, each scored on real GitHub credibility.

Get Vibetest Use →

Browse the catalog