Run Checks - thresholds for failures, flake, coverage, and accessibility outcomes on test runs

Building

Run Checks allow you to define specific standards to be evaluated and attached to your Cypress Cloud runs. This handles a wide range of common use cases, for examople:

  1. A check that fails if your number of failed or flaky tests is higher than a given threshold on your `develop` branch's full regression runs.

  2. An accessibility or UI Coverage check that requires your score to be over a certain percentage on a feature branch

  3. A "max failed test" or "max flaky" test count of zero, with a carve-out for specific specs or folders, to use as quarantine process for problematic tests, without splitting up your test runs.

Having these thresholds available in Cypress Cloud reduces the need for secondary processing steps in other systems and allows features like Webhooks to carry this extra layer of information about a run after it completes. It also creates visibility and consistency about what the current expectations are for "heathy" runs in your team's context, because the rules live at the source of truth for the rest of the run's outcomes.