Skip to main content

What is Watcher?

Watcher is a local-first TestOps workspace by Elixir Nexus for discovering, running, monitoring and reviewing existing Playwright automation projects. You install it as a development dependency in a project you already have, start it from the project root, and work from a dashboard served on your own machine.

Your project remains the source of truth. Watcher reads your package.json scripts and Playwright configuration; it does not ask you to migrate tests, rewrite configuration, or adopt a new project structure.

Community Alpha

This documentation covers Watcher Community Alpha (0.1.0-alpha.1). Windows 11 x64 is the primary Alpha platform; Linux is experimental. See Platform support for the full matrix.

What you can do with Watcher​

CapabilityWhat it gives you
Project discoveryA summary of your project, its Playwright config, test files, package scripts and health checks.
Run PlannerChoose a package script, then adjust projects, workers, retries or headed mode, and preview the exact command before running.
Live monitoringPlanned test count, live progress counters, active tests, ordered console output and a Stop Run control.
Run historyCompleted runs stored locally in .watcher/watcher.db, with per-test final results, failed tests first.
Repeat RunRe-run a historical run with its recorded options, with notices when your scripts or Playwright version changed since.
Reports and artifactsReferences to the HTML, JSON, JUnit, Blob and Allure outputs and attachments a run produced.
Playwright Tools (Experimental)Launch Playwright's own Codegen and UI Mode from the dashboard.

How it works​

  1. Install @elixir-nexus/watcher as a dev dependency in your Playwright project.
  2. Start it with npx watcher start. Watcher discovers the project and serves a dashboard on http://127.0.0.1:5050.
  3. Plan and run a package script from the dashboard. For straightforward playwright test scripts, Watcher runs the project-local Playwright CLI directly and, on Playwright 1.63 or newer, adds its reporter so it can show structured progress.
  4. Review completed runs in Recent Runs, then repeat a run when you need to.

Everything runs on your machine. The dashboard binds to the loopback interface, and Watcher sends no telemetry. Read Local-first, privacy and security for details.

What Watcher is not​

Watcher is not a Playwright replacement, a CI service, a cloud test-management system, or an AI testing platform. You can keep running Playwright directly from the command line and in CI; Watcher's reporter does nothing outside a Watcher-started run.

In this Alpha, Watcher does not provide:

  • Scheduled or recurring runs.
  • Running more than one managed run at the same time.
  • Orchestration of sharded runs across machines. (A script may still pass --shard to Playwright.)
  • Support for test frameworks other than Playwright.
  • Team features such as user accounts, permissions or shared history.
  • An HTML report viewer inside the dashboard.

Next steps​