Conversation automation - fallback to LLM

Configuration Voice Assistant
1

I have the OpenAI (ChatGPT) in place and I have enabled the fallback option: “Prefer handling commands locally”.

Question:
How can I do an automation triggered by a custom sentence to be executed only when a condition is True, else to fallback to LLM (as it would do without the automation) ??

E.g.:

  • Without an automation:
    • I say: “Anna have aples”
    • The local agent “rejects” the sentence.
    • The message goes to LLM/OpenAI/ChatGPT which then is answering … whatever… and it’s ok.
  • With an automation:
    • I say “Anna have aples”
    • The automation is triggered, and:
      • if condition is true, then the list of actions are executed
      • If condition is false, then … how to “reject” the sentence so that it falls back automatically to LLM/OpenAI/ChatGPT ?
description: ""
mode: single
triggers:
  - trigger: conversation
    command: Anna have aples
conditions: []
actions:
  - if:
      - condition: template
        value_template: blabla == True
    then:
      - action: notify.send_message
        metadata: {}
        data:
          message: actions to be executed
      - set_conversation_response: This is the custom response
    else:
      - ????? I want to say that I don't recognize the sentence and fallback to OpenAI, how ... ???

Ps. If I put “nothing” then in case of confition is False then I just get a “Done” answer.