dan/skills

dan d24bedcab3 feat: add playwright-visit skill for browser automation

- visit.py CLI with subcommands: screenshot, text, html, pdf
- Uses system chromium on NixOS (no browser download)
- Fresh profile each run (no cookies/history)
- flake.nix provides playwright devShell
- Options: --wait, --full-page

Useful for JS-heavy sites where WebFetch fails.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

2025-12-29 00:09:50 -05:00

1.6 KiB

Raw Blame History

playwright-visit

Browser automation using Playwright for visiting web pages, taking screenshots, and extracting content.

Overview

Uses headless Chromium to render pages (including JavaScript) and extract content. Useful for:

Screenshots of web pages for visual analysis
Extracting text from JS-heavy sites (where WebFetch fails)
Getting rendered HTML after JavaScript execution
Saving pages as PDF

Prerequisites

NixOS with system chromium (/run/current-system/sw/bin/chromium)
Playwright Python package (provided via flake.nix)

Setup

cd ~/.claude/skills/playwright-visit
nix develop

Usage

# Screenshot
./scripts/visit.py screenshot "https://example.com" /tmp/shot.png
./scripts/visit.py screenshot "https://example.com" /tmp/full.png --full-page

# Extract text
./scripts/visit.py text "https://example.com"

# Get rendered HTML
./scripts/visit.py html "https://example.com"

# Save as PDF
./scripts/visit.py pdf "https://example.com" /tmp/page.pdf

# Wait longer for slow pages (default: 1000ms)
./scripts/visit.py screenshot "https://slow-site.com" /tmp/shot.png --wait 3000

How It Works

Launches headless Chromium using system binary
Creates fresh browser profile (no cookies/logins)
Navigates to URL and waits for network idle
Extracts requested content
Closes browser

Limitations

No authentication support (fresh profile each run)
Requires NixOS with system chromium
Headless only (no visible browser window)

1.6 KiB Raw Blame History