Fume

Fume functions as an autonomous QA tool that generates and maintains Playwright end-to-end tests. It targets software teams that want to automate browser testing without manual script writing, using Loom-style recordings to drive test creation and upkeep. The system emphasizes reducing manual test maintenance by updating selectors and flows as the UI changes, and it supports running tests in the cloud or in a customer’s CI. It also focuses on providing ownership of the Playwright code produced during testing.

Fume analyzes recordings and app behavior to produce Playwright tests, then keeps them healthy through self-healing updates as the app evolves. Tests can run in Fume’s cloud runners or in a customer’s CI, and users retain ownership of the generated code. The service highlights ease of migration from existing tests and emphasizes a lifecycle of recording, generating, running, and self-healing tests.

The tool promotes a fully autonomous QA workflow where no team member must design, write, or maintain tests manually. It supports cloud-based execution and enables integration with existing CI pipelines. It also offers a path to migrate current Playwright/Selenium/Cypress tests and preserve Playwright code ownership.

Fume outlines plans and options such as Team and Enterprise plans, with a focus on cloud runners (including free runners) and CI/CD integration. It discusses ownership of the Playwright output even after cancellation and notes available paths for migrating tests and running in various CI environments.

Use cases

• Migrate existing tests from Playwright, Selenium or Cypress to Fume.
• Describe test goals verbally and have tests generated automatically.