/workspace. A sandbox is replaceable compute that can be recreated without discarding the run’s files.
Continuity model
After a successful turn
Before a response is marked completed, Salambo:- Captures the mutable workspace.
- Saves it through a run-scoped transfer capability.
- Verifies that the recovery point is ready.
- Makes that recovery point authoritative for the run.
Missing sandbox recovery
When a run’s sandbox no longer exists:- Salambo resolves the applied deployment.
- It creates a replacement sandbox.
- It restores the latest ready checkpoint.
- It validates the runtime and workspace.
- It resumes execution on the same run.
Deployment upgrades
WithworkspaceUpgradePolicy: automatic, an existing run moves lazily before its next turn.
The old sandbox remains authoritative until the switch commits. Failed provisioning or restore leaves the old runtime available for a safe retry.
Retention
Salambo currently retains the latest ready checkpoint for recovery. This is continuity infrastructure, not a user-facing checkpoint history or point-in-time restore product.What is checkpointed
| Data | Included |
|---|---|
/workspace | Yes |
| Deployment skills and agent source | No; restored from the immutable deployment |
| Temporary and platform runtime data | No |
| Credentials and control tokens | No |