Meme CLI — scenario audio + test echoes
=======================================

Edit audio here:
  media/audio/scenario-audio.json

Test each scenario (VS Code terminal, Shell Integration on):

  echo "zsh: command not found: foobar"          # command_not_found
  echo "Permission denied"                       # permission_denied
  echo "Could not resolve host: github.com"      # network_error
  echo "Authentication failed"                   # auth_failure
  echo "compilation failed"                      # compile_error
  echo "Tests failed"                            # test_failure
  echo "Build failed"                            # build_failure
  echo "Merge conflict in file.js"               # git_conflict
  echo "No such file or directory"               # file_not_found
  echo "WARNING: something looks odd"            # warning
  echo "Downloading package... 45% complete"     # download_progress
  echo "Installing dependencies"                 # install_progress
  echo "Compiling sources"                       # compile_progress
  echo "Waiting for input"                       # waiting_input
  echo "Already up to date"                      # git_success
  echo "Listening on http://localhost:3000"      # server_ready
  echo "Port 8080 is already in use"             # server_error
  echo "Application is ready"                    # app_ready
  echo "All tests passed"                        # test_success
  echo "BUILD SUCCESS"                           # build_success

Wait ~1 second between commands.
Empty audio path => speech fallback.
