Asi playwright

Browser automation via Playwright MCP. Use for web scraping, taking screenshots, interacting with web pages, testing web UIs, and automating browser tasks. Headless browser support.

install
source · Clone the upstream repo
git clone https://github.com/plurigrid/asi
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/plurigrid/asi "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/playwright" ~/.claude/skills/plurigrid-asi-playwright-0f4605 && rm -rf "$T"
manifest: skills/playwright/SKILL.md
source content

Playwright Browser Automation

Control browsers via Playwright MCP server.

When to Use

  • Web scraping and data extraction
  • Taking screenshots of web pages
  • Interacting with web UIs (clicking, typing, navigating)
  • Testing web applications
  • Automating browser-based workflows
  • Filling forms and submitting data

Setup

MCP server configured in

~/.mcp.json
:

{
  "playwright": {
    "command": "npx",
    "args": ["-y", "@anthropic-ai/mcp-server-playwright"]
  }
}

Common Tools

Navigation

  • navigate_page
    - Go to a URL
  • new_page
    - Open new browser tab
  • list_pages
    - Show open pages

Interaction

  • click
    - Click elements
  • fill
    - Type into input fields
  • select
    - Choose from dropdowns
  • press
    - Press keyboard keys

Capture

  • take_screenshot
    - Screenshot current page
  • get_page_content
    - Get page HTML
  • get_text
    - Extract visible text

Evaluation

  • evaluate
    - Run JavaScript in page context

Example Workflows

Screenshot a Page

  1. navigate_page(url="https://example.com")
  2. take_screenshot()

Fill a Form

  1. navigate_page(url="https://example.com/form")
  2. fill(selector="#email", value="user@example.com")
  3. fill(selector="#password", value="secret")
  4. click(selector="button[type=submit]")

Extract Data

  1. navigate_page(url="https://example.com/data")
  2. get_text(selector=".results")

Tips

  • Use CSS selectors or XPath for element targeting
  • Wait for page loads before interacting
  • Browser runs headless by default
  • Screenshots are useful for debugging