A workflow to cut loops and API costs in n8n multi-agent setups

This Reddit post describes a workflow for running several AI agents together in n8n. The author says it addresses agents looping endlessly, producing malformed JSON, and quickly spending Claude API budget. The workflow uses a PFAF structure with separate roles, budget tracking, and JSON validation.

Key points

  • The post focuses on a multi-agent workflow in n8n.
  • It targets endless loops, malformed JSON, and fast Claude API spending.
  • PFAF separates agents into defined roles with strict rules.
  • The setup includes budget tracking and JSON validation.
  • The main lesson is to control agent behavior before token costs grow.

Quick term guide

AI agents
AI agents are AI tools that can carry out steps toward a goal, not just answer once.
Claude API
A way for software or automations to call Claude AI directly.
validation
Checking whether real people understand, want, or would use an idea before spending more time on it.
multi-agent workflow
A setup where two or more AI tools each handle different parts of a job and pass results between them
multi-agent
A setup where several AI agents each handle a different subtask and work together to complete a larger goal.
agent workflow
A set of steps an AI follows automatically to complete a series of tasks in order.
token costs
Token costs are the fees paid for the text an AI model reads and writes.
token cost
The money or usage spent when sending text to an AI model and getting text back.
Read original