Skip to content

Test AutomationSimplified view

NL2Test, from a sentence to a Playwright test

How NL2Test turns a plain-language description into an executable Playwright test, with a model planning the test and an MCP server driving a real page.

  • Playwright
  • Model Context Protocol
  • VS Code
  • OpenAI

10 stepsPress play, or use next and previous to follow the flow one step at a time.

All steps

What this shows

NL2Test turns a natural-language description of a scenario into an executable Playwright test. This diagram follows one request through the pieces involved: the editor, the Model Context Protocol (MCP) client and server, the model that plans the test, and the browser tools that act on a real page.

Design notes

  • The model plans; the tools act. The model decides what to do, and the MCP server exposes the tools that do it, so the model never needs direct access to a browser.
  • Ground the test in the real page. Reading a snapshot of the page before choosing locators keeps the test tied to what is actually on screen.
  • A person stays in the loop. The result is a test a developer reviews and runs, not one that is trusted blindly.

This is a simplified view of the flow, drawn to explain the idea. The NL2Test project page has the project itself.

More workflows

AI and Agents

How an MCP request flows

A walk through one Model Context Protocol request, from the client through authentication, memory and the model to a server that coordinates tools.

8 steps 8 parts Simplified view

AI and Agents

Multi-agent supervisor

A supervisor agent receives a request, hands sub-tasks to specialist agents one at a time, keeps shared memory, and combines their results into one answer.

8 steps 7 parts Simplified view

AI and Agents

RAG pipeline

How retrieval-augmented generation works end to end, from loading documents into a vector store to answering a question with the closest passages.

8 steps 7 parts Simplified view