
Shortest
Open-source QA via natural language — write `shortest("Log in")` and AI drives Playwright.
About
Shortest is an open-source framework from Antiwork that lets you write end-to-end tests in natural language on top of Playwright. A test is literally shortest("Log in with email and password") — the AI receives an accessibility snapshot of the page and drives the browser with element refs instead of fragile selectors. It supports multiple LLM providers (GLM, Azure, DashScope, Siliconflow), runs headless or headed, and integrates with GitHub Actions.
The honest picture: as a young open-source project, docs and community support are thinner than commercial rivals, and you bring your own LLM API key (so test runs cost model tokens). Action replay caching is disabled in snapshot mode since refs aren't stable across runs, which can slow suites down. Still, for teams that want AI-driven E2E without a vendor, it's the most credible free option in the category.
What it can do
- Natural-language test definitions via `shortest()`
- Accessibility-snapshot driven browser automation
- Multiple LLM provider support
- GitHub Actions integration
- Headless and headed Playwright execution
Setup
Install with npm i @antiwork/shortest, add a shortest.config.ts with your base URL and LLM provider, then run npx shortest.
Prerequisites
Node.js; an LLM API key (GLM, Azure, DashScope, or Siliconflow); a web app to test.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 49; Trending score 68/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.