---
title: "Discussion on \"The winning setup used less than half the tokens and finished more tasks\" | Hashnode"
description: "Discussion on \"The winning setup used less than half the tokens and finished more tasks\". IBM Research compared two ways of handing an agent what it has already learned.\nAgentic Context Engineering injects the whole playbook every time.\nALTK-Evolve retrieves only the guidelines that fit th"
url: "https://hashnode.com/posts/the-winning-setup-used-less-than-half-the-tokens-and-finished-more-tasks/6a8156cd577346708b85b979"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [The winning setup used less than half the tokens and finished more tasks](https://blog.bhuveshdhiman.com/the-winning-setup-used-less-than-half-the-tokens-and-finished-more-tasks?utm_source=hashnode&utm_medium=feed)

IBM Research compared two ways of handing an agent what it has already learned. Agentic Context Engineering injects the whole playbook every time. ALTK-Evolve retrieves only the guidelines that fit th

Aug 16 · 2 min read · blog.bhuveshdhiman.com  
Read the article

-   [#ai](https://hashnode.com/tag/ai)
-   [#llm](https://hashnode.com/tag/llm)
-   [#context-engineering](https://hashnode.com/tag/context-engineering)

## Written by

[Bhuvesh Dhiman](https://hashnode.com/@bhuveshdhiman)

@bhuveshdhiman · Mumbai, India

  
[View profile](https://hashnode.com/@bhuveshdhiman)

Sharing learnings as a senior software engineer

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
