AI by Hand ✍️

AI by Hand ✍️

Verbatim Tail

Keeping the recent rounds whole

Prof. Tom Yeh's avatar
Prof. Tom Yeh
Aug 19, 2026
∙ Paid

Library › Context Problems

  1. The Context Window

  2. Window Occupancy

  3. Growth per Turn

  4. Remaining Space

  5. Turn Budget

  6. Retrieval Footprint

  7. Retrieval per Turn

  8. Search Rounds

  9. Window Sizing

  10. Truncation

  11. Compaction

  12. The Compaction Bill

  13. Compacted Retrieval

  14. The Compaction Threshold

  15. Verbatim Tail

  16. Stateless API

  17. Retrieval Cost

  18. Context Tax

  19. Quadratic Cost

  20. Cost Forecast

Compaction does not summarize everything behind the system prompt. It leaves the most recent rounds alone. An agent that summarized the round it was in the middle of would lose the thread it is holding, so the summary stops short and the live end of the conversation stays verbatim. That leaves the window in three parts: the prompt, the summary of what is old, and the rounds still in their own words.

Paid members: the worksheet and its printable PDF are below ↓

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2026 Tom Yeh · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture