> For the complete documentation index, see [llms.txt](https://cogna8-io.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cogna8-io.gitbook.io/docs/core-concepts/enforcement-boundary.md).

# Enforcement Boundary

Cogna8 is strongest when its boundary is clear.

***

## Inside the Boundary

{% hint style="success" %}
**Cogna8 verifies and enforces:**

* Whether the state an action depends on is consistent, current, and in scope
* Whether conflicts are surfaced explicitly
* Whether required state is present and resolved before an action proceeds
* Whether trace and audit records connect every decision to the state that produced it
* Whether scope boundaries prevent cross-context contamination
  {% endhint %}

## Outside the Boundary

{% hint style="info" %}
**Cogna8 does not govern:**

* Base model behavior or language generation quality
* Agent personality, tone, or conversational style
* General workflow orchestration and routing design
* External storage architecture
* UI presentation outside integrity semantics
  {% endhint %}

***

## The Principle

> **Cogna8 verifies whether the system should act on state - not how the model should sound while talking about it.**

{% hint style="success" %}
**When each layer does its job, the whole stack becomes more trustworthy than any single layer can achieve alone.** Action authority is the missing piece that connects memory (what the system knows) to action (what the system does) through verified, traceable, scope-aware trust.
{% endhint %}