How to Automatically Retry Flaky Tests with GitHub Actions

Retrying only the tests that failed turns a transient flake into a pass without re-running the whole suite.

You can retry at the test level with jest.retryTimes(2), or run the suite, then re-run failures with jest --onlyFailures. Reserve retries for genuinely flaky tests, since blanket retries hide real regressions.

Steps

• Add jest.retryTimes(2, { logErrorsBeforeRetry: true }) in a setup file for in-suite retries.
• Or run the suite once and re-run only failures with jest --onlyFailures.
• Treat a test that needs retries as a bug to fix, not a permanent crutch.

Workflow

jobs:
  test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 20
          cache: npm
      - run: npm ci
      - name: Run tests, retrying failures once
        run: npx jest || npx jest --onlyFailures

Gotchas

• --onlyFailures needs Jest's cache, so do not clear it between the two runs.
• Auto-retry masks flakiness; track which tests retry so you can fix root causes.
• Latchkey managed runners also auto-heal transient infrastructure failures, which is a different layer from test-level retries.

Related guides