Execute Command Action

A GitHub Action that executes an arbitrary command and captures its output to files, including stdout, stderr, and exit code. The action streams command output in real-time and forwards signals to the command.

Breaking Changes

BREAKING CHANGE: This action now writes stdout and stderr to files instead of GitHub Action outputs. This change addresses issues with passing large outputs to subsequent steps (which could cause "Argument list too long" errors). The outputs stdout and stderr have been replaced with stdout_file and stderr_file, which contain paths to the files where the output is stored.

Features

• Execute any single command
• Capture standard output and standard error to temporary files
• Output file paths available as action outputs
• Stream output in real-time to the workflow logs
• Optionally hide outputs from the workflow log to protect sensitive data
• Forward signals (SIGINT, SIGTERM, SIGQUIT, SIGHUP, SIGPIPE, SIGABRT) to the running command
• Commands are executed directly without a shell (no shell operators like |, &&, >)

IMPORTANT: This action executes commands directly without a shell. This means shell features like pipes (|), redirects (>), command chaining (&&, ||), and glob expansion (*) are not available.

Usage

steps:
  - name: Checkout
    uses: actions/checkout@v6

  - name: Execute Command
    id: exec
    uses: retailnext/exec-action@main
    with:
      command: 'echo "Hello World"'

  - name: Print Output
    run: |
      echo "Exit Code: ${{ steps.exec.outputs.exit_code }}"
      echo "Stdout File: ${{ steps.exec.outputs.stdout_file }}"
      echo "Stderr File: ${{ steps.exec.outputs.stderr_file }}"
      echo "Stdout Content:"
      cat "${{ steps.exec.outputs.stdout_file }}"
      echo "Stderr Content:"
      cat "${{ steps.exec.outputs.stderr_file }}"

Inputs

command

Required The command to execute with its arguments.

The command is executed directly without a shell. Executables in your PATH can be used without specifying the full path (e.g., npm, ls, git).

success_exit_codes

Optional Exit codes that should be treated as success. Can be individual codes (e.g., "0,1,2") or ranges (e.g., "0-2,5,10-15"). Default is "0".

This is useful when your command may exit with non-zero codes that should still be considered successful. For example, some linters return specific exit codes for warnings vs errors, or you may want to accept multiple exit codes as valid outcomes.

hide_outputs

Optional When set to "true", the stdout and stderr from the command are only written to the respective output files and are not written to the stdout/stderr of the action itself. Default is "false".

This is useful for commands that emit sensitive data (e.g., secrets, tokens) to their output. The outputs are hidden from the GitHub Actions log but remain available to subsequent workflow steps via the stdout_file and stderr_file output paths.

Outputs

stdout_file

Path to the file containing the standard output of the executed command. The file is located in the directory specified by the RUNNER_TEMP environment variable.

stderr_file

Path to the file containing the standard error of the executed command. The file is located in the directory specified by the RUNNER_TEMP environment variable.

exit_code

The exit code of the executed command (as a string).

Examples

Run a build command

- name: Build Project
  id: build
  uses: retailnext/exec-action@main
  with:
    command: 'npm run build'

- name: Check Build Status
  if: steps.build.outputs.exit_code == '0'
  run: echo "Build succeeded!"

Handle errors

- name: Run Command
  id: run
  uses: retailnext/exec-action@main
  with:
    command: 'some-command-that-might-fail'
  continue-on-error: true

- name: Handle Failure
  if: steps.run.outputs.exit_code != '0'
  run: |
    echo "Command failed with exit code ${{ steps.run.outputs.exit_code }}"
    echo "Error output:"
    cat "${{ steps.run.outputs.stderr_file }}"

Accept multiple exit codes as success

- name: Run Linter
  uses: retailnext/exec-action@main
  with:
    command: 'eslint .'
    # Treat exit codes 0 (no issues) and 1 (warnings only) as success
    success_exit_codes: '0,1'

Accept a range of exit codes

- name: Run Tests
  uses: retailnext/exec-action@main
  with:
    command: 'pytest'
    # Treat exit codes 0-5 as success
    success_exit_codes: '0-5'

Mix individual codes and ranges

- name: Complex Command
  uses: retailnext/exec-action@main
  with:
    command: 'some-tool --check'
    # Accept 0, any code from 10-15, and 20 as success
    success_exit_codes: '0,10-15,20'

Hide sensitive outputs from the log

- name: Fetch Secret Data
  id: fetch
  uses: retailnext/exec-action@main
  with:
    command: 'my-tool --output-secret'
    hide_outputs: 'true'

- name: Process Secret Data
  run: |
    # The output is hidden from the log but available via the file path
    process-data < "${{ steps.fetch.outputs.stdout_file }}"