Outputs

Output	Example	Description
filemoverexpress-version	1.1.1	Resolved version, without a v prefix.
filemoverexpress-path	/opt/hostedtoolcache/filemoverexpress/1.1.1/x64/filemoverexpress	Absolute path to the CLI binary.
cache-hit	false	true when the CLI came from the runner tool cache instead of a download.
config-dir	/home/runner/work/_temp/filemoverexpress	Directory exported as FME_CONFIG_DIR.
config-file	/home/runner/work/_temp/filemoverexpress/configuration.yaml	<config-dir>/configuration.yaml.
transfer-profile	default	Name of the transfer profile the action wrote. Empty when bucket isn't set.

Using outputs

Give the step an id and read outputs from steps.<id>.outputs:

- id: fme
  uses: lowlysre/setup-filemoverexpress@v1
  with:
    filemoverexpress-version: '1.x'
    bucket: my-media-bucket
    region: us-east-1

- name: Show the generated config
  shell: bash
  env:
    FME_VERSION: ${{ steps.fme.outputs.filemoverexpress-version }}
    FME_CACHE_HIT: ${{ steps.fme.outputs.cache-hit }}
    FME_CONFIG_FILE: ${{ steps.fme.outputs.config-file }}
  run: |
    echo "Version: ${FME_VERSION}"
    echo "Cache hit: ${FME_CACHE_HIT}"
    cat "${FME_CONFIG_FILE}"

- name: Upload with the profile the action wrote
  shell: bash
  env:
    FME_PROFILE: ${{ steps.fme.outputs.transfer-profile }}
  run: filemoverexpress upload "${FME_PROFILE}" ./renders

Passing outputs through env instead of writing expressions directly inside run keeps values out of the script text, which is the pattern linters such as zizmor expect.