antiwork/shortest

QA via natural language AI tests

What it solves

Shortest simplifies end-to-end (E2E) testing by allowing developers to write test cases in natural language instead of complex, brittle scripts. It reduces the effort required to maintain tests as the UI changes, as the AI handles the execution logic.

How it works

Built on top of Playwright, the framework uses the Anthropic Claude API to interpret natural language descriptions of test steps and execute them in the browser. It can handle complex flows like GitHub 2FA logins and API endpoint validation. Developers can also integrate custom callback functions for database assertions or use lifecycle hooks to manage test state.

Who it’s for

Software engineers and QA testers who want to speed up the creation of E2E tests and reduce the maintenance overhead associated with traditional selector-based testing.

Highlights

  • Natural Language Execution: Write tests as simple strings (e.g., "user can login with email and password").
  • AI-Powered: Leverages Claude to drive browser interactions.
  • Flexible Chaining: Supports sequential test chains and reusable test flows.
  • Extensible: Includes .after() callbacks for custom logic and standard lifecycle hooks (beforeAll, beforeEach, etc.).
  • API Testing: Supports testing API endpoints using natural language descriptions.
  • Integration Support: Built-in support for GitHub 2FA and Mailosaur for email validation.

Related

  • Project
    alumnium-hq/alumnium

    An AI-native library and MCP for end-to-end testing that enables natural language interactions and assertions using Appium, Playwright, or Selenium.

  • Project
    ai-dashboad/flutter-skill

    An AI-native E2E testing tool that lets AI agents control and test apps across 10 platforms using natural language and the Model Context Protocol (MCP).

  • Project
    browserbase/skills

    A collection of skills for Claude Code that enable browser automation, web research, and AI-powered UI testing via Browserbase.

  • Project
    bug0inc/passmark

    An open-source Playwright library for AI-powered browser regression testing that uses natural language steps and multi-model consensus assertions to prevent test fragility.

  • Project
    test-zeus-ai/testzeus-hercules

    An open-source AI testing agent that converts plain-English Gherkin steps into automated end-to-end web tests, removing the need for manual coding.