Introduction
What is NZT
NZT is a working method for coding agents. It gives the model shape through method, memory and brakes so it performs steadily, from idea to product.
NZT does not add knowledge to the model, and it does not limit it either. It gives it shape —method, memory and brakes— so it performs steadily instead of in bursts.
It covers the whole path: understand what the user needs, define it, design it, build it, verify it and ship it.
Where the name comes from
It comes from Limitless, the film where NZT is a fictional pill. It teaches Eddie nothing new: it gives him ordered access to what he already knows.
The bad consequences in the film came from taking it without structure. Here, the structure is the product: stopping points, persisted state and human review.
NZT by the numbers
| Number | What it counts |
|---|---|
110 |
installed skills |
1 |
specific leaf loaded per task |
0 |
subagents, plugins or MCP |
~180 |
kernel lines always present |
47 |
installer checks |
Which agents it works with
NZT installs into Claude Code and Codex CLI. Once installed, there are no new commands to learn: you open your agent and ask for what you need in plain language.