I Built an AI Agent that Builds Teams of Agents in n8n (free template)

I Built an AI Agent that Builds Teams of Agents in n8n (free template)

by Nate Herk

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQMMachine learning
🎙 Nate Herk 👥 964K 📅 June 18, 2025 ⏱ 19 min 👁 85K 📄 tutorial 🧭 2026-08-28
Available in: English (current) Français

Keywords

n8nAI agentworkflow generationClaude Opus 4no-code

Summary

The video presents a practical tutorial on building an AI agent within n8n that can autonomously generate complete n8n workflows from natural language requests. The creator demonstrates three live examples: a Slack-based assistant with calendar and Gmail tools, a personalized email workflow with HubSpot integration, and a Google Sheets outreach system. The core mechanism involves a main agent that delegates to a ‘developer agent’ powered by Claude Opus 4 with thinking enabled, which outputs a JSON workflow structure. This JSON is then submitted to the n8n API to create a new workflow. The creator emphasizes the simplicity of the approach, using a single Google Doc as context, and provides a free template for viewers. The video also covers setup steps, pricing considerations, and the creator’s iterative development process, including token optimization. The overall tone is enthusiastic, highlighting the potential of AI to democratize workflow automation.

145 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video offers substantial practical value by providing a working, downloadable template and a clear, step-by-step explanation of the system’s architecture. The demonstrations are convincing and show real, functional outputs. The argumentation is solid, grounded in the creator’s direct experience and iterative testing. The creator transparently discusses limitations, such as occasional node name errors and the need for manual tweaks, which strengthens credibility. The approach is innovative in its minimalism, using a single example document rather than a large knowledge base, and the reasoning behind this choice is well-articulated.

Scientific Rigor, Source Quality, Title Accuracy

The video is a tutorial based on the creator’s own work, so it does not cite external scientific sources. The main reference is the n8n platform itself, and the creator mentions using Claude Opus 4 from Anthropic. The description provides links to the creator’s community and tools, which are relevant for obtaining the template. The title accurately reflects the content, and the video’s structure with clear chapters aids comprehension. The methodology is reproducible, but the lack of external validation or comparison with other approaches limits the scientific rigor.

192 words

Title / Content Match

The title accurately reflects the content: the video demonstrates building an AI agent in n8n that can generate other n8n workflows.

Quality & Reliability

7/10

The video provides a detailed, reproducible tutorial with live demonstrations and a free template. The methodology is transparent, though the approach relies on a single example and lacks formal validation or comparison with alternative methods.

Chapters

Cited Sources

Concurring Sources

  • n8n documentation — Official documentation for n8n, which supports the technical details of the workflow creation.
  • Anthropic Claude — The AI model used, confirming its capabilities for JSON generation and thinking.

Contribution & Novelties

The video presents a novel, minimal-resource approach to AI-driven workflow generation in n8n, using a single example document instead of a large knowledge base. This reduces cost and complexity while still producing functional workflows. The creator’s method of feeding a Google Doc as context and using Claude Opus 4 with thinking is a practical innovation. The free template and detailed walkthrough make the technique accessible to a wide audience.

Pour aller plus loin :

  • n8n documentation — Official documentation for n8n, useful for understanding node structures and API.
  • Anthropic Claude — The AI model used in the video, with details on capabilities and pricing.
  • JSON — The data format used for workflow representation, fundamental to the approach.

117 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quantity and technical level, reflecting the tutorial's practical nature. The lower score in information quality and reliability suggests that while the content is useful, it lacks formal validation and external references.

Reliability 7/10

💬 Très positif. Sur les 30 commentaires analysés, l'enthousiasme est unanime, avec des éloges sur la clarté, la valeur pratique et l'innovation de l'approche.