Outputs

The action exposes the CLI’s exit code and combined output so subsequent workflow steps can react to either.

Output	Description
exit_code	The CLI’s exit code as a string. 0 means the registry passed validation; non-zero indicates findings or a runtime error.
output	The full combined stdout + stderr produced by the CLI. The same text the action posts as a PR comment when enabled.

Job exit status

If the CLI exits non-zero, the action’s final step re-exits with the same code, so the job fails. There is no need to read exit_code to fail the build — that is the default behaviour. Use exit_code only when you want to conditionally continue past a failure (for example, to upload an artifact before the job dies).

Reading outputs

- id: pm
  uses: pullminder/action@v1
  with:
    command: validate

- name: Save raw CLI output as artifact
  if: always()
  run: |
    cat <<'OUT' > pullminder-output.txt
    ${{ steps.pm.outputs.output }}
    OUT
- uses: actions/upload-artifact@v4
  if: always()
  with:
    name: pullminder-output
    path: pullminder-output.txt

if: always() is required to capture output on a failed validation, since the CLI exit status fails the job by default.

Reading the exit code

- id: pm
  uses: pullminder/action@v1
  continue-on-error: true
  with:
    command: validate

- name: Open issue on failure
  if: steps.pm.outputs.exit_code != '0'
  run: gh issue create --title "Registry validation failed" --body "$OUTPUT"
  env:
    GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
    OUTPUT: ${{ steps.pm.outputs.output }}

continue-on-error: true is required because the action re-exits with the CLI status code; without it, the workflow would short-circuit before the follow-up step runs.