I’m trying to build an n8n workflow to generate a very long “Bed-Time History” narration script (totaling 15,000-16,500 words). The script needs to be divided into 22 numbered sections, each 1,000-1,100 words.
I’m using the Google Gemini Chat Model via the Basic LLM Chain node and have a specific two-step workflow:
- Step 1 (Outline): I provide a topic, and the first LLM generates a 22-bullet outline for the script. After this, there’s an instruction: “STOP. Wait for my ‘CONTINUE’.”
- Step 2 (Full Script Generation): Once I issue the “CONTINUE” command, the LLM must generate ALL 22 script sections consecutively (starting with “Section 1”, “Section 2”, etc.), following the outline it previously created.
How can I do that?