General-purpose Playwright automation for coding agents
An Agent Skill that enables coding agents to write and execute Playwright automation on the fly, from simple page tests to complex multi-step flows. It is also packaged as a Claude Code Plugin for convenient installation.
Claude autonomously decides when to use this skill based on your browser automation needs, loading only the minimal information required for your specific task.
Made using Claude Code.
- Any Automation Task - Claude writes custom code for your specific request, not limited to pre-built scripts
- Visible Browser by Default - See automation in real-time with
headless: false - Portable executor - Runs file and inline scripts with stable module resolution
- Progressive Disclosure - Concise SKILL.md with full API reference loaded only when needed
- Safe Cleanup - Smart temp file management without race conditions
- Comprehensive Helpers - Optional utility functions for common tasks
This repository contains a standard Agent Skill and a Claude Code plugin wrapper. The recommended installation method is Vercel's skills CLI, which installs skills into the native locations for supported agents.
Use this skill when the agent needs to write a real Playwright program: loops, assertions, multiple contexts, network interception, screenshots, video, or a script you want to keep and rerun. It also provides dev-server detection and a small set of focused helpers.
For straightforward interactive browsing, start with Microsoft's official
@playwright/cli and install its
agent skills with playwright-cli install --skills. For tool-based browser
control with accessibility snapshots, use
playwright-mcp. This project is
the code-first option when the automation itself is the useful artifact.
This repository uses the plugin format with a nested structure:
playwright-skill/ # Plugin root
├── .claude-plugin/ # Plugin metadata
└── skills/
└── playwright-skill/ # The actual skill
└── SKILL.md
The repository keeps the skill inside the plugin's skills/ directory. Installers handle that layout automatically; manual copying is only a fallback for clients without an installer.
Install globally for your user:
npx skills add lackeyjb/playwright-skill --skill playwright-skill --global --yesInstall only for the current project by omitting --global:
npx skills add lackeyjb/playwright-skill --skill playwright-skill --yesTo target specific agents, add --agent followed by one or more agent IDs:
npx skills add lackeyjb/playwright-skill --skill playwright-skill --agent claude-code cursor --global --yesAfter installation, run setup from the installed skill directory:
npm run setupSee the skills CLI documentation for supported agents and options.
Install via Claude Code's plugin system for automatic updates and team distribution:
# Add this repository as a marketplace
/plugin marketplace add lackeyjb/playwright-skill
# Install the plugin
/plugin install playwright-skill@playwright-skill
# Navigate to the skill directory and run setup
cd ~/.claude/plugins/marketplaces/playwright-skill/skills/playwright-skill
npm run setupVerify installation by running /help to confirm the skill is available.
Agent Skills are supported by Claude Code, Cursor, GitHub Copilot, Codex,
Gemini CLI, OpenCode, and other clients. Install the directory containing
SKILL.md using the client’s documented skill path. If a client has no
installer, copy skills/playwright-skill/ into its documented skill directory
and run npm run setup there.
- Download and extract the latest release from GitHub Releases
- Copy only the
skills/playwright-skill/folder to:- Global:
~/.claude/skills/playwright-skill - Project:
/path/to/your/project/.claude/skills/playwright-skill
- Global:
- Navigate to the skill directory and run setup:
cd ~/.claude/skills/playwright-skill # or your project path npm run setup
Run /help to confirm the skill is loaded, then ask Claude to perform a simple browser task like "Test if google.com loads".
After installation, ask your agent to test or automate a browser task. It will write custom Playwright code, execute it, and return results with screenshots and console output.
"Test the homepage"
"Check if the contact form works"
"Verify the signup flow"
"Take screenshots of the dashboard in mobile and desktop"
"Test responsive design across different viewports"
"Fill out the registration form and submit it"
"Click through the main navigation"
"Test the search functionality"
"Check for broken links"
"Verify all images load"
"Test form validation"
- Describe what you want to test or automate
- Your agent writes custom Playwright code for the task
- The universal executor (run.js) runs it with proper module resolution
- Browser opens (visible by default) and automation executes
- Results are displayed with console output and screenshots
Default settings:
- Headless:
false(browser visible unless explicitly requested otherwise) - Slow Motion:
0msby default; setSLOW_MOwhen useful - Screenshots: Helper screenshots use the OS temp directory; set
PW_ARTIFACT_DIRto choose another location
playwright-skill/
├── .claude-plugin/
│ ├── plugin.json # Plugin metadata for distribution
│ └── marketplace.json # Marketplace configuration
├── skills/
│ └── playwright-skill/ # The actual skill (Claude discovers this)
│ ├── SKILL.md # What Claude reads
│ ├── run.js # Universal executor (proper module resolution)
│ ├── package.json # Dependencies & setup scripts
│ └── lib/
│ └── helpers.js # Optional utility functions
│ └── API_REFERENCE.md # Full Playwright API reference
├── README.md # This file - user documentation
├── CONTRIBUTING.md # Contribution guidelines
└── LICENSE # MIT License
Claude will automatically load API_REFERENCE.md when needed for comprehensive documentation on selectors, network interception, authentication, visual regression testing, mobile emulation, performance testing, and debugging.
- Node.js
- Playwright (installed via
npm run setup) - Chromium (installed via
npm run setup)
Playwright not installed?
Navigate to the skill directory and run npm run setup.
Module not found errors?
Ensure automation runs via run.js, which handles module resolution.
Browser doesn't open?
Verify headless: false is set. The skill defaults to visible browser unless headless mode is requested.
Install all browsers?
Run npm run install-all-browsers from the skill directory.
Agent Skills are folders of instructions, scripts, and resources that agents can discover and use to do things more accurately and efficiently. When you ask Claude to test a webpage or automate browser interactions, Claude discovers this skill, loads the necessary instructions, executes custom Playwright code, and returns results with screenshots and console output.
This Playwright skill implements the open Agent Skills specification, making it compatible across agent platforms.
Contributions are welcome. Fork the repository, create a feature branch, make your changes, and submit a pull request. See CONTRIBUTING.md for details.
- Agent Skills Specification - Open specification for agent skills
- Claude Code Skills Documentation
- Claude Code Plugins Documentation
- Plugin Marketplaces
- API_REFERENCE.md - Full Playwright documentation
- GitHub Issues
MIT License - see LICENSE file for details.