https://medium.com/better-ml/how-to-prompt-a-capable-llm-to-almost-gold-at-imo-2025-39b69e91d296

It shows context engineering plus pipeline engineering with off the shelf models can achieve the same level capabilities of super models.

It also shows reasoning can be decomposed to several simple steps.


<
Previous Post
Simplify your Agent 'vibe building' flow with ADK and Gemini CLI
>
Next Post
Interesting behavior of agents