browsinglisted

# Browser Automation with agent-browser ## Overview Use `agent-browser` to control Chromium for web automation tasks. It is a CLI tool that manages the full browser lifecycle — no external browser process needed. **Announce:** "I'm using the browsing skill with agent-browser to automate Chrome." ## When to Use - Navigating websites and extracting content - Filling and submitting forms - Taking screenshots of pages or elements - Multi-step web workflows - Acceptance testing of web applications ## Quick Reference ```bash # Navigate and extract page content agent-browser run "navigate https://example.com; extract text" # Take a screenshot agent-browser run "navigate https://example.com; screenshot /tmp/page.png" # Click an element agent-browser run "navigate https://example.com; click '#submit-button'" # Fill a login form agent-browser run "navigate https://example.com/login; type '#email' 'user@example.com'; type '#password' 'secret'; click 'button[type=submit]'" ``` ## Commands ### Navigation ```bash # Navigate to URL agent-browser run "navigate https://example.com" # Wait for element before proceeding agent-browser run "navigate https://example.com; wait-for '.content-loaded'" # Wait for text to appear agent-browser run "navigate https://example.com; wait-text 'Welcome'" ``` ### Content Extraction ```bash # Extract page as markdown (best for reading content) agent-browser run "navigate https://example.com; extract markdown" # Extract plain text agent-browser