What are the technical prerequisites?

You need Bun installed for validation, the Agent tool enabled, and three Playwright MCP instances configured in your .mcp.json file.

How does it handle dynamic test data?

It uses 'testdata:' directives within Gherkin files to trigger project-specific setup scripts, capturing JSON output to provide context and credentials for the test runner.

Why does this skill require three Playwright instances?

The skill uses three parallel MCP server instances (playwright-1, playwright-2, playwright-3) to run multiple browser tests concurrently, significantly speeding up the testing process.

What is the browser-test skill for Claude Code?

It is a specialized skill that allows Claude to manage the entire QA lifecycle, from writing Gherkin test specs to executing them via Playwright and reporting bugs.

Can it fix broken tests automatically?

Yes, in Phase 2b (Spec Repair), the skill analyzes failures to determine if the test is stale or if it found a bug, offering to update the spec if the change was intentional.

Playwright Gherkin Browser Testing

Name: Playwright Gherkin Browser Testing
Author: robmadole

byrobmadole

0•

Security & Testing

Orchestrates end-to-end QA browser testing workflows using Gherkin specifications and parallel Playwright execution.

This skill provides a complete QA automation engine for Claude Code, enabling the generation, execution, and maintenance of browser-based tests. It analyzes pull requests or feature descriptions to create Gherkin feature files, then dispatches them to three concurrent Playwright MCP instances for high-speed parallel execution. Beyond simple testing, it includes an intelligent 'Spec Repair' phase that can distinguish between legitimate application changes and actual bugs, automatically updating test specs when needed and providing an audited report of all findings.

Key Features

01Detailed QA reporting with categorized results and automated auditing

02Parallel test execution using three concurrent Playwright MCP subagents

03Intelligent spec repair and gap analysis to maintain test suite health

040 GitHub stars

05Automated Gherkin spec generation from PR diffs, features, or text descriptions

06Dynamic test data injection using custom testdata directives

Use Cases

01Running existing end-to-end test suites in parallel to reduce development feedback loops

02Updating legacy test specifications automatically when application behavior evolves

03Verifying UI changes in a Pull Request by automatically generating and running relevant specs

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add robmadole/agent-tools browser-test

For use in Claude.ai and ChatGPT