Data retention
Context enforces no retention policy: data is deleted only explicitly, resource by resource, by its owners.
Where each kind of data is stored and how it is encrypted: Data at rest.
How retention works
- In a deployment that runs in your own cluster, data stays in your own database and file storage.
- Your platform owners govern retention: admins delete content on your organization’s own schedule.
How long each kind of content is kept
| Content | Kept until |
|---|---|
| Files and deliverables | Deleted by its owner, then kept in Trash until emptied. On organizations that use the newer permissions system, Trash is purged 30 days after an item is deleted. |
| Conversations and run history | Deleted with the conversation |
| Agent memory | Deleted item by item |
| Workspaces | Deleted by the owner or an admin |
Content is deleted only by its owners or an admin. Apart from the Trash purge above, nothing is removed on a schedule.
Legal hold
Legal hold is not a product feature. Under a hold, admins leave the content in place; your own preservation and eDiscovery tools can reach the database and storage directly.