Playwright CLI QA Workflows

These specs are lightweight, readable automation scripts for Playwright CLI. Each spec is a Markdown file with runnable playwright-cli code blocks.

Running a spec 

scripts/playwright-run-spec.sh docs/specs/admin-smoke.md

The runner executes fenced code blocks marked with playwright-cli and ignores all other content.

Spec format

  • Use a fenced code block starting with ` ```playwright-cli`.
  • Use step "<label>" to name the next CLI action.
  • Use cli <command> ... to invoke Playwright CLI commands.
  • Use open_browser to open the browser with BASE_URL and optional args.

Each step label applies to the next cli call.

Example 

step "Open admin"
open_browser

step "Check dashboard"
cli run-code "$(cat <<'EOF'
async page => {
  await page.getByRole('heading', { name: 'Dashboard', level: 2 }).waitFor();
}
EOF
)"

Environment variables

  • BASE_URL (default http://localhost:8081)
  • PLAYWRIGHT_OPEN_ARGS (example: --headed)
  • ARTIFACT_ROOT (default artifacts/playwright)
  • FLOW_NAME (defaults to spec filename without .md)
  • PLAYWRIGHT_CLI_PKG (default @playwright/cli@0.1.0)
  • PLAYWRIGHT_CLI_CMD (optional override path to a CLI binary)

Artifacts and a report.txt file are written under:

artifacts/playwright/<flow>__<timestamp>/

Notes

  • The function passed to run-code runs in the Playwright (Node.js) context and controls the page object. Use page.evaluate(...) to run code in the browser context, and pass values via shell variables or function arguments.