Web Access & Browser Automation

0•

Automates browser navigation, web searches, and data extraction across authenticated and dynamically rendered web pages using Chrome DevTools Protocol (CDP).

Web Access equips Claude Code with human-like browsing intelligence and robust automation capabilities using a multi-tiered approach: WebSearch, WebFetch, Jina AI markdown conversion, and direct Chrome DevTools Protocol (CDP) proxy connections. By attaching seamlessly to your running Chromium browser (Chrome or Edge), it leverages existing login sessions to navigate dynamic Single Page Applications (SPAs), scrape protected platforms like Xiaohongshu or WeChat, inspect deep DOM structures, capture screenshots, and coordinate parallel sub-agent research—all while operating safely in isolated background tabs.

Key Features

01Connects directly to active Chromium browsers via CDP proxy to preserve authenticated user sessions
02Implements adaptive multi-tier retrieval using WebSearch, WebFetch, Jina AI, or full browser emulation
03Extracts structured media, captures video frames, and bypasses complex dynamic rendering barriers
040 GitHub stars
05Executes full DOM evaluation, mouse clicks, keyboard input, and scroll-based lazy loading
06Coordinates isolated background tabs for parallel, sub-agent web research and token conservation

Use Cases

01Conducting automated parallel research across multiple independent web sources simultaneously
02Scraping dynamic content and media from authenticated platforms, private dashboards, or social media sites
03Verifying claims, policies, and official documentation directly from first-hand source websites