playwright-scenarios¶
Claude Code plugin for scenario-driven Playwright testing — crawl, record, evaluate, convert, review, and generate tests from human-readable scenario markdown.
What is it?¶
playwright-scenarios is a Claude Code plugin that lets you author browser-driven test scenarios as flat markdown files, audit those scenarios against the live site, and translate them into test classes. The plugin ships 9 commands, 6 skills, and 13 extended tags.
playwright-scenarios works with any project that supports Playwright for browser automation. The default test generation stack is Kotlin + Kotest StringSpec + Playwright-for-Java, with support for TypeScript and Python planned.
How it works¶
graph TD
A[Your website]
A -->|"/crawl-site"| B[Scenario]
A -->|"/record-scenario"| B
A -->|"/doc-to-scenarios"| B
B -->|"/review-scenario"| C[Reviewed<br/>scenario]
C -->|"/scenario-to-tests"| D[Test suite]
Three quick-start authoring paths converge on /scenario-to-tests. See the Workflow page for a fourth path (LLM-authored documents, which routes through /doc-to-scenarios) and a side-by-side comparison.
A scenario is an LLM-optimized markdown representation of the flow you want to test — readable by humans, mechanically processable by Claude. You produce one by recording a browser session, crawling the site, or converting an existing document — each writes to its own folder (scenarios/crawl/, scenarios/record/, scenarios/convert/). Hand-edit or delete in place if you want; then /review-scenario audits against the live site, and /scenario-to-tests turns the reviewed scenarios into a runnable test suite at <test_dir>/<command>/<scenario-name>/.
Learn more¶
-
Step-by-step walkthrough from installation to generated tests.
-
Definitions for scenario, test case, source folder, fixture, tag, and more.
-
The full pipeline from document to generated tests, with decision points.
-
What the plugin can test, what it can't, and where gaps live instead.
-
Guidance for writing test documents that convert cleanly into scenarios.