FAQ¶

Conceptual and scope questions about the playwright-scenarios plugin. For runtime errors and "I tried X and got Y" debugging, see Troubleshooting instead.

Do I have to use the bookstore demo from the tutorial?¶

No — the bookstore demo is just a concrete target for the tutorial. Any URL Claude can reach via Playwright is fair game: your local dev server, a staging environment, a public site. Wherever the tutorial says http://localhost:8080, substitute your own URL. You can skip the docker run step in Step 1 entirely if you already have a server running.

The tutorial calls out every substitution under a For your project: note. The Bring your own site preamble at the top of the tutorial summarizes what's swappable.

Why do I need Node.js?¶

For playwright-cli, the binary that /review-scenario, /scenario-to-tests, and /crawl-site shell out to during their live-site exploration phase. It's distributed as an npm package (@playwright/cli), and npm requires Node.js. The dependency chain:

/review-scenario, /scenario-to-tests, /crawl-site
      ↓ invoke
playwright-cli skill
      ↓ wraps
@playwright/cli (npm package)
      ↓ installed via
npm install -g @playwright/cli@latest
      ↓ requires
Node.js

/record-scenario is the only authoring command that doesn't need playwright-cli — it launches Playwright codegen via the Gradle recordScenario task in your project. If you only ever record, Node.js is skippable.

What if my project isn't Kotlin?¶

/scenario-to-tests currently generates only Kotlin + Kotest StringSpec tests with the Playwright-for-Java bindings. Python and TypeScript generators are planned but not yet shipped.

The rest of the pipeline is language-agnostic — /crawl-site, /record-scenario, /doc-to-scenarios, and /review-scenario all produce or operate on plain markdown scenarios. You can author scenarios for any web app today; you just can't auto-generate the tests for non-Kotlin stacks yet. When the matching generator lands, set test_language / test_framework in .claude/playwright-scenarios.local.md and re-run /scenario-to-tests against your existing scenarios.

Which command should I use to author a scenario?¶

Starting point	Use	Writes to
You have a written document (test plan, requirements doc, meeting notes, ACs)	`/doc-to-scenarios <path>`	`<scenario_dir>/convert/`
You know the flow but have no document	`/record-scenario [url]`	`<scenario_dir>/record/`
You don't know what flows exist yet	`/crawl-site <start-url>`	`<scenario_dir>/crawl/`

All three feed the same downstream pipeline: /review-scenario to audit against the live site, then /scenario-to-tests to generate executable tests. For the full decision tree, see Workflow.