chrome-devtools-mcp and web-eval-agent
These are complements—Chrome DevTools provides the low-level browser automation primitives (protocol interaction, DOM inspection, script execution) that a web evaluation agent would build upon to autonomously test and assess web applications.
About chrome-devtools-mcp
ChromeDevTools/chrome-devtools-mcp
Chrome DevTools for coding agents
Implements a Model-Context-Protocol (MCP) server that bridges AI coding agents with Chrome's DevTools Protocol, enabling performance tracing, network inspection, and automated browser actions through Puppeteer. Supports both full-featured and lightweight ("slim") modes, integrating with agents like Claude, Gemini, and Cursor for debugging, screenshot capture, and source-mapped stack trace analysis.
About web-eval-agent
refreshdotdev/web-eval-agent
An MCP server that autonomously evaluates web applications.
Leverages Playwright for browser automation and integrates with the operative.sh backend to enable LLM agents (in Cursor, Cline, Windsurf) to autonomously test web apps end-to-end. Captures network traffic, console logs, and screenshots, returning structured evaluation reports that agents use to validate their own code changes. Includes stateful browser setup for authenticated workflows via cookie/localStorage persistence.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work