saving-unique-id-to-cache-workflow.yml

name: My workflow

on: [pull_request]

jobs:
  build_id:
    steps:
      - name: 'Set build id'
        id: build_id
        # add a step output `steps.build_id.outputs.id` as the unique id
        run: echo "::set-output name=id::$(date +%s)"
      - name: 'Save build id'
        # store the build id to a folder that we can later cache.
        run: |
          mkdir -p .build-id 
          echo $BUILD_ID > .build-id/id
        env:
          BUILD_ID: ${{ steps.build_id.outputs.id }}
  tests:
    # require the build to run after the build_id job
    needs: build_id
    name: Cypress run
    runs-on: ubuntu-latest
    strategy:
      matrix:
        # run 2 copies of the current job in parallel
        containers: [1, 2]
    steps:
      - name: Checkout
        uses: actions/checkout@v1
      # because of "record" and "parallel" parameters
      # these containers will load balance all found tests among themselves
      - name: Cypress run
        uses: cypress-io/github-action@v1
        with:
          record: true
          parallel: true
        env:
          # pass the Dashboard record key as an environment variable
          CYPRESS_RECORD_KEY: ${{ secrets.CYPRESS_RECORD_KEY }}