I see abstract AI agent architectures everywhere. But no one explains how to build them in practice. Here's a practical guide to doing it with n8n: 1. 𝐒𝐢𝐧𝐠𝐥𝐞 𝐚𝐠𝐞𝐧𝐭𝐬 Selected variants: • Using tools • Mixing tools with MCP servers • With a router (a fancy name for a condition) • With a human in the loop (Slack approval) • Dynamically calling other agents 2. 𝐌𝐮𝐥𝐭𝐢𝐩𝐥𝐞 𝐚𝐠𝐞𝐧𝐭𝐬 Selected variants: • Working sequentially • Hierarchy with parallel execution and shared tools • Hierarchy with a loop and shared RAG 3. 𝐁𝐞𝐬𝐭 𝐩𝐫𝐚𝐜𝐭𝐢𝐜𝐞𝐬 Here’s what works best for me: 1. Ask the agent to plan its work, pursue the plan until the objective is met, and reflect after each iteration. 2. Add memory so the agent can track its progress. 3. Use a loop to better control complex processes. 4. Suggest common tool usage patterns in the prompt (e.g., the order). 5. Make sure tools and MCP servers have clear descriptions. 6. Check “Return Intermediate Steps” in the Agent settings to debug the thought process. 7. Select “Error Workflow” in the workflow settings to handle exceptions. 8. If you're using the community version without global variables, create a dedicated workflow to get values by variable name instead of hardcoding them. 9. Clearly assign roles and objectives (e.g., planner, researcher, reviewer). Learn by building, not theorizing. 🎁 You can download my poster as an n8n workflow definition (json, Google Drive): https://lnkd.in/dG5ciYAc Hope that helps! -- P.S. Enjoy this? You might also like "J.A.R.V.I.S. for PMs: How to automate anything with n8n and MCP:" https://lnkd.in/dSW6_KiT | 20 comments on LinkedIn