web-eval-agent
Verified Safeby withRefresh
Overview
An autonomous agent that executes and debugs web applications by navigating, capturing network traffic, collecting console errors, and providing UX reports directly in a code editor.
Installation
python mcp_server.pyEnvironment Variables
- OPERATIVE_API_KEY
- USE_LOCAL_BACKEND
Security Notes
The project leverages Playwright for browser automation, which inherently involves powerful actions like JavaScript injection (via `page.evaluate`) and DOM manipulation (e.g., `agent_overlay.js` patching `elementFromPoint`). It uses a Flask-SocketIO server with `cors_allowed_origins='*'` (though bound to localhost, mitigating some direct cross-origin risks) and `allow_unsafe_werkzeug=True`, which are permissive settings generally intended for development and not ideal for distribution. Standard output and error streams are redirected to `/dev/null`, which can obscure critical information or runtime errors. The agent transmits the `OPERATIVE_API_KEY` to an external `operative-backend.onrender.com` for LLM calls and validation, necessitating trust in this third-party service.
Similar Servers
chrome-devtools-mcp
Control and inspect a live Chrome browser programmatically via an MCP server, enabling AI coding agents to perform reliable automation, in-depth debugging, and performance analysis.
brightdata-mcp
Enables AI agents to access, search, extract, and navigate the live web in real-time without being blocked.
mcp
This server provides Hyperbrowser's Model Context Protocol (MCP) interface, offering tools for web scraping, structured data extraction, crawling, and general-purpose browser automation using AI agents like OpenAI's CUA and Anthropic's Claude Computer Use.
rocketship
This MCP server assists AI coding agents by providing Rocketship testing examples, optimal test structure, schema information, CLI guidance, and codebase analysis to help agents write better Rocketship tests.