Skip to main content

testivai env

Print this machine's part of the capture environment: what every capture records about where it came from, so you can see what your baselines will be compared against, and why two machines render the same page differently.

npx testivai env            # human-readable
npx testivai env --fonts    # also fingerprint the installed fonts (fc-list)
npx testivai env --json     # machine-readable
TestivAI capture environment (this machine)
  os:        linux 6.8.0-1017-azure x64 (ubuntu 24.04)
  ci:        github-actions (runner image ubuntu24/20251012.1.0)
  container: no
  label:     none (set TESTIVAI_ENV_LABEL to name this environment)
  fonts:     412 installed, fingerprint 9326f323d545
  policy:    environmentPolicy "warn"
  • os: the platform, release and architecture; on Linux, the distro from /etc/os-release. The distro, not the kernel, decides the fonts and the text renderer.
  • ci: the CI provider, and the runner image on GitHub Actions.
  • container: whether this runs inside a container (/.dockerenv or /run/.containerenv).
  • label: TESTIVAI_ENV_LABEL, a name you give this environment (for example the container image tag). A changed label counts as a rendering change.
  • fonts (--fonts): how many fonts fc-list lists and a short hash of the list. Two machines with different fingerprints draw text differently. Not recorded per capture, since listing fonts costs a process.
  • policy: environmentPolicy from .testivai/config.json (warn, separate or strict; see how it works).

The browser, framework, adapter, capture method and render settings depend on the browser a test opens, so each capture records them in its own metadata.json (see the Extension API).