playwright-automation

name	playwright-automation
description	Browser automation templates for pi-scheduler using Playwright
version	1.0.0

Playwright Automation

pi-scheduler includes three built-in templates for browser automation via Playwright. These templates execute as JavaScript scripts in a child_process, keeping pi-scheduler-core zero-dependency.

Prerequisites

Install Playwright in your project directory (the cwd you set on the automation):

npm install playwright
npx playwright install chromium

Built-in Templates

web-screenshot

Navigates to a URL and saves a screenshot.png in the automation's cwd.

scheduler.instantiateTemplate('web-screenshot', {
  name: 'Dashboard screenshot',
  cwd: 'D:/repos/myproject',
  params: { url: 'http://localhost:3000' },
});

Params: url — target URL (no query strings; use simple URLs like http://host:port/path).

---

url-health-check

Loads a URL with a real browser and exits with code 1 if HTTP status ≥ 400. Useful for SPAs or pages behind auth that need JavaScript to render.

scheduler.instantiateTemplate('url-health-check', {
  name: 'API health',
  intervalMinutes: 5,
  cwd: 'D:/repos/myproject',
  params: { url: 'http://localhost:8080/health' },
});

Params: url — endpoint to check.

---

login-flow

Submits a login form and reports the resulting page title. Useful for verifying an auth flow is still working.

scheduler.instantiateTemplate('login-flow', {
  name: 'Auth smoke check',
  intervalMinutes: 30,
  cwd: 'D:/repos/myproject',
  params: { url: 'https://myapp.internal/login' },
});

Params: url — login page URL.
Env vars: PW_USERNAME, PW_PASSWORD — credentials (never put secrets in the params object).

The login-flow template uses heuristic selectors ([name="username"], [type="submit"]). Copy the script to a custom template and adjust selectors for your specific app.

---

Custom Playwright Templates

Register your own template at runtime:

scheduler.registerTemplate({
  id: 'my-playwright-task',
  name: 'My browser task',
  description: 'Custom Playwright automation.',
  defaultInterval: 60,
  scriptType: 'javascript',
  command: null,
  script: `
const { createRequire } = require('node:module');
const req = createRequire(process.cwd() + '/package.json');
const { chromium } = req('playwright');
(async () => {
  const browser = await chromium.launch({ headless: true });
  const page = await browser.newPage();
  await page.goto('https://yourapp.com');
  // ... your automation logic ...
  await browser.close();
})();
  `.trim(),
  subagentConfig: null,
  requiredParams: [],
});

The key pattern is createRequire(process.cwd() + '/package.json') — this resolves playwright from the node_modules of the automation's cwd, so each project can have its own Playwright version.

Running the Integration Test

cd D:/repos/pi-scheduler
npm install playwright
npx playwright install chromium
node test-integration/playwright.mjs

The test is skipped gracefully if Playwright is not installed.