# Context management

> Context management is the daily practice of selecting, scoping, refreshing, and auditing the context an AI agent receives.

It is about the right context, not maximum context.

Ambience's view: for teams, context management means turning useful work into source-linked memories, narrowing access with scopes, redacting before storage, and reviewing retrieval health before agents rely on the context.

## Related

- [Context management for AI agents](https://ambience.sh/company-context/context-management-for-ai-agents)
- [Company Context Map](https://ambience.sh/company-context/company-context-map)
- [ContextOps](https://ambience.sh/company-context/contextops)