Introduction
Every organization has rules. Most keep them in a wiki nobody reads, written for humans and ignored by the automation that does the work. An AI-native org does the opposite: it keeps its rules in one file, in the root of the codebase, addressed to both the people and the agents that change it - a small, non-negotiable floor that holds whether the next change came from a person or a model. Cogitave is one org that works this way, and this module uses its floor as the worked example.
The file is AGENTS.md, an open standard for exactly this purpose: every coding and automation agent - and every human in a session - reads the nearest AGENTS.md, and the root file is the floor that all sub-trees inherit. Cogitave's own root AGENTS.md is its managed policy; it opens with a section titled "Non-negotiable rules (the floor)" and lists seven of them. This module teaches those seven as a concrete floor you can learn from and adapt to your own estate.
IMPORTANT
A floor is the floor, not a suggestion. Cogitave's seven rules are numbered 1 to 7 in its AGENTS.md, and this module teaches them in that order. When the floor and a habit disagree, the floor wins.
Why a floor at all
In an AI-native org, agents change the same code and infrastructure that people do. A floor that both a person and an agent can read is what lets a narrow, cheap model act unattended and still stay correct - the environment carries the rules, so they hold whether the author was human or not. In Cogitave's estate you are working inside a control plane, and every change is auditable evidence: a reviewer or an auditor will read your diff.
A good floor also stays short and delegates the detail to an owning standard. Cogitave's does exactly this: AGENTS.md states the rule and points you at the canonical document under standards/docs/standards/. Learn the floor from the one file; learn the specifics from the standard it cites. Never paraphrase a rule from memory when the source is one link away.
What you will get from this module
Not a tour of opinions - a working command of the seven rules on Cogitave's floor, a clear picture of what enforces each one (a hook, a CI gate, an org ruleset, or policy-as-code), and the ability to apply a floor like it to your very first change.