~/posts
7 entries · newest first · try ls posts in the shell
- 2026-07-11 How agents remember
Memory is not a feature you add to an agent — it is a decision about what the agent is.
- 2026-07-11 Introducing melchizedek
A capable model is not a capable system — coordination is what turns talent into predictable work.
- 2026-07-11 One agent, seven windows
How a market-analysis agent synthesizes many narrow data functions into a single defensible view.
- 2026-07-11 The patient advocate
A lesson in building an AI second mind — one that recalls, connects, and researches so a patient can understand their diagnosis and ask the right questions.
- 2026-07-11 The spec before the image
Predictable image generation comes from approving a JSON contract, not from polishing a prompt — reverse-engineered from one real print.
- 2026-07-11 Welcome to the Lyceum
Why this site exists, what it will become, and how to get around it.
- 2026-07-11 What is an agent?
A working definition, built the Socratic way — with an interactive dialogue instead of a diagram.