Agent skill

web-browser

Allows to interact with web pages by performing actions such as clicking buttons, filling out forms, and navigating links. It works by remote controlling Google Chrome or Chromium browsers using the Chrome DevTools Protocol (CDP). When Claude needs to browse the web, it can use this skill to do so.

View SKILL.md on GitHub Repository
Stars 3
Forks 1

Install this agent skill to your Project

npx add-skill https://github.com/MykalMachon/llm-config/tree/main/skills/web-browser

SKILL.md

Web Browser Skill

Minimal CDP tools for collaborative site exploration.

Start Chrome

bash
./tools/start.js              # Fresh profile
./tools/start.js --profile    # Copy your profile (cookies, logins)

Start Chrome on :9222 with remote debugging. Auto-detects platform (macOS/Linux/WSL) and searches for Chrome/Chromium. On WSL, automatically adds --no-sandbox flag.

Navigate

bash
./tools/nav.js <https://example.com>
./tools/nav.js <https://example.com> --new

Navigate current tab or open new tab.

Evaluate JavaScript

bash
./tools/eval.js 'document.title'
./tools/eval.js 'document.querySelectorAll("a").length'
./tools/eval.js 'JSON.stringify(Array.from(document.querySelectorAll("a")).map(a => ({ text: a.textContent.trim(), href: a.href })).filter(link => !link.href.startsWith("https://")))'

Execute JavaScript in active tab (async context). Be careful with string escaping, best to use single quotes.

Screenshot

bash
./tools/screenshot.js

Screenshot current viewport, returns temp file path

Pick Elements

bash
./tools/pick.js "Click the submit button"

Interactive element picker. Click to select, Cmd/Ctrl+Click for multi-select, Enter to finish.

Didn't find tool you were looking for?

Be as detailed as possible for better results