wico — Playwright Agent Skills

Reusable Playwright E2E testing knowledge for AI coding assistants. This project provides a CLI that scaffolds agent skills into any Playwright project, formatted for the platform you use — Claude Code, Cursor, GitHub Copilot, or a generic .agent-skills/ directory.

Project Description

This project extracts the generic, battle-tested patterns from real-world Playwright E2E suites into a standalone, installable package. The idea is simple: instead of every QA engineer teaching their AI assistant the same Playwright best practices from scratch, you install a set of curated skills once and your agent immediately knows how to write, debug, and review tests properly.

The skills cover:

• Playwright API patterns (waitForResponse, toPass, expect.poll, network-first safeguards)
• Test data strategy (static vs dynamic factories)
• Page Object Model conventions (selectors, composition, naming)
• Project conventions (MUST/SHOULD/WON'T rules, file organization, CI/CD)
• Test debugging (failure patterns, root cause classification, app bug vs test bug decision tree)
• Test generation and planning (templates, checklists, exploration workflow)
• Playwright CLI reference (browser automation for interactive exploration)

The CLI detects your project setup, asks which platforms and skill packs you want, replaces template placeholders with your project info, and generates the files in the right format for each platform.

How to Install

Prerequisites

• Node.js (>= v18) — Download from nodejs.org

node -v   # must be >= v18

Quick Start

Navigate to your Playwright project and run:

npx wico-playwright-agent-skills init

That's it. The CLI will detect your project, ask which platforms and skills you want, and generate the files.

Development Setup

If you want to contribute or run from source:

# Clone the repository
git clone [email protected]:willcoliveira/qualiow-playwright-skills.git
cd qualiow-playwright-skills

# Install dependencies
npm install

How to Use the CLI

The CLI walks you through 5 steps:

  wico — Playwright Agent Skills v0.1.0

  Step 1: Project Detection
  ─────────────────────────
  ✓ Found playwright.config.ts
  ✓ TypeScript project detected

  Step 2: Agent Platform(s)
  ─────────────────────────
  Which AI assistant(s) do you use? (space to select)
  > [x] Claude Code        → .claude/skills/
    [x] Cursor              → .cursor/rules/
    [ ] GitHub Copilot      → .github/copilot-instructions.md
    [x] Generic             → .agent-skills/

  Step 3: Skill Packs
  ─────────────────────────
  > [x] Core patterns (playwright-patterns, data-strategy, test-review)
    [x] Playwright CLI reference
    [x] Project templates (conventions, POM, debugging, generation, planning)

  Step 4: Project Info (for templates)
  ─────────────────────────
  Project name: my-e2e-suite
  Base URL: https://staging.example.com
  Fixture import path: ../fixtures/test-fixture (or "none")
  Page objects dir: src/pages
  Test dir pattern: src/tests

  Step 5: Confirm & Generate
  ─────────────────────────
  Will create 24 files across 3 platforms.
  Proceed? (Y/n)

  ✓ Done! Next: customize <!-- YOUR PROJECT: ... --> markers

Running from Source (Development)

If you cloned the repo and want to test locally:

# Run with tsx (no build needed)
npx tsx bin/init.ts init

# Or build first, then run the compiled version
npm run build
node dist/bin/init.js init

Type Checking

npm run lint

Building for Distribution

npm run build

What Gets Generated

Platform Output Formats

Each platform has its own convention for where AI instructions live. The CLI generates files in the right format for each one:

Skill Packs

Core Patterns (shipped as-is)

These are generic Playwright best practices that apply to any project:

Playwright CLI Reference (shipped as-is)

Complete reference for the playwright-cli browser automation tool:

Project Templates (customized with your project info)

These templates get {{PLACEHOLDER}} values replaced with your project-specific info during generation. They also include <!-- YOUR PROJECT: ... --> markers where you should add your own details after setup.

Template Placeholder System

Templates use a simple {{PLACEHOLDER}} syntax with a lightweight regex engine (no Handlebars dependency).

String Placeholders

Conditional Blocks

{{#if HAS_CUSTOM_FIXTURE}}
import { test } from '{{FIXTURE_IMPORT_PATH}}'
{{/if}}
{{#if NO_CUSTOM_FIXTURE}}
import { test } from '@playwright/test'
{{/if}}

Extension Points

After generation, search for <!-- YOUR PROJECT: ... --> markers in the generated files. These are placeholders where you should add your project-specific details:

<!-- YOUR PROJECT: Add your component inventory here -->
<!-- YOUR PROJECT: Document your test data files here -->
<!-- YOUR PROJECT: List your required environment variables here -->

After Setup

Once the files are generated, here's what to do next:

1. Review the generated files for your platform and make sure they look right
2. Search for <!-- YOUR PROJECT: ... --> markers and fill in your project-specific details (components, fixtures, env vars, tags, etc.)
3. Customize the MUST/SHOULD/WON'T rules in project-conventions.md to match your team's agreements
4. Add your failure patterns to test-debugging.md — the common issues your team hits
5. Document your page objects and fixtures in test-generation.md so the agent knows what's available
6. Test it — ask your AI assistant to write a test or debug a failure and see if it follows the skills

Project Structure

qualiow-playwright-skills/
├── bin/
│   └── init.ts                       # CLI entry point
├── src/
│   ├── cli.ts                        # CLI orchestrator (5-step flow with @clack/prompts)
│   ├── prompts.ts                    # Project detection (playwright.config.ts, tsconfig.json)
│   ├── generator.ts                  # File generation orchestrator
│   ├── template-engine.ts            # {{PLACEHOLDER}} + {{#if}} replacement engine
│   └── platforms/
│       ├── claude.ts                 # .claude/skills/ generator
│       ├── cursor.ts                 # .cursor/rules/*.mdc generator
│       ├── copilot.ts                # .github/copilot-instructions.md generator
│       └── generic.ts                # .agent-skills/ generator
├── skills/
│   ├── core/                         # Shipped as-is (generic Playwright knowledge)
│   │   ├── playwright-patterns.md
│   │   ├── data-strategy.md
│   │   └── test-review.md
│   ├── templates/                    # Shipped with {{PLACEHOLDERS}} for customization
│   │   ├── page-object-conventions.md
│   │   ├── project-conventions.md
│   │   ├── test-debugging.md
│   │   ├── test-generation.md
│   │   └── test-planning.md
│   ├── playwright-cli/               # Playwright CLI skill (as-is)
│   │   ├── SKILL.md
│   │   └── references/
│   │       ├── request-mocking.md
│   │       ├── running-code.md
│   │       ├── session-management.md
│   │       ├── storage-state.md
│   │       ├── test-generation.md
│   │       ├── tracing.md
│   │       └── video-recording.md
│   └── indexes/                      # Platform-specific SKILL.md indexes
│       ├── skill-index.md            # Generic index with decision tree
│       ├── claude-skill.md           # Claude SKILL.md format
│       ├── cursor-rules.mdc          # Cursor frontmatter format
│       └── copilot-instructions.md   # Copilot single-file format
├── examples/                         # Example generated output per platform
│   ├── claude/
│   ├── cursor/
│   ├── copilot/
│   └── generic/
├── package.json
├── tsconfig.json
├── tsup.config.ts                    # Build config (single ESM bundle)
├── .gitignore
└── LICENSE

Tech Stack

Contributing

If you'd like to contribute new skills, improve existing templates, or add support for another platform:

1. Fork the repository
2. Create a feature branch (git checkout -b feat/my-skill)
3. Make your changes
4. Run npm run lint to verify types
5. Submit a pull request

License

MIT