Skip to main content

Going further

You've got a Playwright dashboard up. A few directions to go from here.

More than test runs​

The test_runs data source is just one preset. Define more sources for anything else you want on the same dashboard:

  • Deploys — deployed_at, git_sha, environment, actor. Status grid by environment.
  • Errors — Sentry / Bugsnag webhooks. Number-with-delta for "errors today".
  • Latency — APM timeseries via webhook. Line chart with bucketed avg.

The flow is identical: define a DataSource with a schema_definition, give it a slug, point a webhook (or tiler import-*) at /ingest/<slug>, and add panels that reference it.

Multiple environments​

Run one tiler server per environment (tiler-staging, tiler-prod) or filter on a environment column you carry through every record. The status grid widget doubles as an "environment overview" if you set group_column: "environment".

TV mode polish​

Press the monitor icon in the editor (or set settings.tv_mode: true via the API). The chrome disappears and the grid fills the viewport. Combine with [data-theme="dark"] for OLED-friendly displays.

The WebSocket subscription means the TV updates as records land — no manual refresh.

Auth​

The editor is read-write by default. For shared environments:

tiler.config.ts
auth: {
  basic: {
    user: "admin",
    pass: process.env.TILER_PASS,
  },
  webhookSecret: process.env.TILER_WEBHOOK_SECRET,
},

Or replace the basic-auth check entirely with a custom authorize callback (SSO, JWT, whatever). See packages/server/README.md for the contract.

Where things live​

Get help​