Instead of hardcoding prompts inside a workflow, you can load them dynamically from an external source that is easy to edit and audit. In this episode I use Google Sheets as a prompt configuration store for an n8n AI Agent.
How the workflow works
- A Chat node takes the user's input.
- A Google Sheets node (Get Rows) fetches the system prompt from a spreadsheet.
- The AI Agent node combines the fetched system message with the user prompt and calls OpenAI.
Why parameterize prompts externally
- Google Sheets keeps an automatic change history — you can audit how the prompt evolved.
- Clients or teammates can edit prompts without access to the n8n workflow.
- Configuration is centralized, which prevents copy-paste drift across workflows.
In the video I demo two examples: a history assistant answering questions about a given date, and an item-valuation agent returning min/mid/max prices as JSON — switching between them only by editing one spreadsheet cell.
Watch the episode
The video is available on YouTube: youtube.com/watch?v=4utnYrpX4xw.
All episodes are collected on the Automation Academy page (in Polish). Subscribe to my YouTube channel so you do not miss new material.
The video is in Polish; this post summarizes what it covers.
Comments