New: Cookbooks and AI ExplanationsStep-by-Step recipes to solve problems connected to Roadmaps and Cheat Sheets. Need more details? Use AI buttons for structured and simple explanations with concrete examples throughout the whole platform.Take a look
Everyone reaches for an agent by default, and half the time a plain script would work better and cost less.
What you'll have at the end
A written call, workflow or agent, for one real task, backed by a traced coverage number and the four-question test, with the winning architecture's exact step split.
You need
A specific LLM feature you're about to build or already run as a rough prototype, with a sense of the operations it needs to perform.
Not covered
Picking which named agent framework to build with, once you've already decided the task needs one.
Decide between LangGraph, CrewAI, and a plain agent loop
Once this recipe says the task is genuinely open-ended, that's where you pick which framework runs the agent.
Should you build your own agent, or buy one?
If you've concluded the task needs an agent, this is the next question: build it yourself or buy one.
Checked 15 Aug 2026
Part of the AI Engineering cookbook