Hey everyone,
One thing I keep seeing people struggling with prompts inside the AI Agent node, agents that call tools when they shouldn’t, return broken JSON, or fail silently on edge cases.
Generic prompt guides don’t help here because they’re not written for n8n’s execution environment.
So I put together a vault of 50 prompts built specifically for n8n: system prompts for common agent types, tool-use instructions, strict output formatting, error handling, and security guards (prompt injection, system prompt leakage, persona hijacking).
First 5 community users get the PDF for free, use discount code N8N at checkout !
Happy to answer questions here if anything needs more context.