Development · GitHub repo
vibetest-use
MCP server running parallel browser agents for automated website QA testing
vibetest-use is an MCP server that uses Browser-Use and Playwright to launch multiple AI agents that crawl local or live websites. It identifies UI bugs, 404 errors, dead links, and accessibility issues directly inside AI coding tools like Claude Code and Cursor. This enables developers to run an automated test-fix-retest cycle without leaving their coding assistant.
What it's for
- Run automated browser QA tests on localhost development servers
- Detect broken links, 404s, and dead buttons across a website
- Verify visual fixes and regressions directly through Claude Code or Cursor
Features
- Model Context Protocol (MCP) server integration for Cursor and Claude Code
- Multi-agent parallel testing with configurable agent counts
- Supports both headless and visible browser testing via Playwright Chromium
- Works on both localhost environments and public live websites
- Powered by Gemini 2.0 Flash for automated browser navigation
Pricing
Open source. Free and open-source (requires user's Google Gemini API key)
Details
- Type
- GitHub repo
- Category
- Development
- Topic
- Testing & QA
- Open source
- Yes
- Platforms
- Command line, Self-hosted
- For
- Developers building web applications using AI coding assistants like Cursor and Claude Code
- Alternative to
- Cypress, Playwright
- Tags
- mcp, browser use, qa, testing, claude code, cursor, playwright
Why it was saved
- @browser_use: One-line automated website QA directly inside an AI coding agent; the video shows it finding broken CV/GitHub links and verifying the fix with a second run.
Compare
Related
Agent-first control plane backed by 2,400+ curated AI developer skills
Agent harness optimization system with skills, hooks, and configs for coding agents
Command-line browser automation tool optimized for AI coding agents and testing.
Browser automation CLI and engine purpose-built for AI agents
Portable agent skills that enforce design quality in AI-generated frontends
Open directory and CLI for discovering and installing AI agent skills
Updated