> 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/reference/glossary.md).

# Glossary

| Term                       | Definition                                                                                                                                           |
| -------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- |
| **State Item**             | The atomic unit of governed information in Cogna8. A typed, keyed, lifecycle-tracked object.                                                         |
| **Canonical Key**          | A unique identifier that maps different phrasings of the same concept to a single key for comparison.                                                |
| **Governed State**         | Information that participates in integrity checks, policy enforcement, and action gating.                                                            |
| **Memory**                 | Retained context that does not currently participate in action integrity. May be promoted.                                                           |
| **Conflict**               | A detected contradiction between two or more active state items sharing the same canonical key.                                                      |
| **Action Gate**            | The deterministic authorization engine that evaluates whether an action is authorized to proceed (allow/warn/block) and issues a structured receipt. |
| **Action Integrity Score** | A compact advisory signal summarizing action readiness derived from gate checks.                                                                     |
| **Policy**                 | A defined rule that specifies required conditions for an action type to proceed.                                                                     |
| **Scope**                  | The context boundary within which state is visible and governing (thread/entity/global).                                                             |
| **Scope Bleed**            | When state from one context silently influences actions in another context.                                                                          |
| **State Promotion**        | The transition from remembered information to governed state.                                                                                        |
| **Supersession**           | When a state item is replaced by a newer value. The old item is preserved in audit.                                                                  |
| **Trace**                  | The structured record connecting a decision to the specific state and checks that produced it.                                                       |
| **Why Panel**              | The inspection surface for reviewing the causal chain behind any governance outcome.                                                                 |
| **Pipeline**               | The multi-stage processing path from input through extraction, conflict detection, gating, and commit.                                               |
| **Decay**                  | The mechanism by which governed state loses active participation over time based on relevance.                                                       |
| **Evidence Link**          | The connection between a governed state item and the source message that introduced it.                                                              |
| **Snapshot**               | A point-in-time capture of the governed state picture for review and comparison.                                                                     |
| **Integrity**              | The property that information an agent acts on is current, consistent, correctly scoped, and policy-compliant.                                       |
| **Authority Runtime**      | The system that sits between agent intent and real-world side effects, deciding whether state mutations and external actions are authorized.         |
| **Authorization Receipt**  | A structured record proving what state was relied on, what policies were evaluated, and why an action was authorized or blocked.                     |
| **Authorization Evidence** | The inspection surface showing the causal chain behind any authorization decision.                                                                   |