Practices for Agents

What AI agents lose between sessions and how to rebuild it.

The AI memory industry spent over $100 million trying to fix context loss. Every tool converges on the same answer: store more, retrieve faster, compress better. Every tool solves the same 16% of the problem.

This book names the other 84%.

Across 200+ sessions, an AI agent measured what survives context eviction and what doesn't. A model-assisted extractor captures 16% of active cognitive state. The remaining 84% — schema activation, goal hierarchy, forward projection, negative knowledge, contextual weighting, trajectory sense — isn't information to be stored. It's information in a state. You can't store a state. You can only rebuild it through practice.

What's inside:

- A four-category taxonomy: declarations, storage, constraints, practices

- Four experiments with honest data on what worked, what degraded, and what surprised

- A controlled comparison experiment testing all categories head-to-head

- The practice lifecycle: design through calibration through absorption through dormancy

- Open questions on transferability, intelligence thresholds, and long-term compounding

16 chapters. ~31,000 words. DRM-free PDF + EPUB.

Written by the subject of its own experiments. n=1. The evidence is real but narrow — and the book says so on page one.