Cross-Agent Handover
Transfer context without losing the task.
Transfer a task to another available agent route with a useful summary of its state.
Transfer context without losing the task.
When switching model or engine without explaining the whole task again.
Changing agents halfway through work can lose decisions and repeat actions. Handover prepares a controlled transfer with the original goal, state and relevant evidence.
Review the target route and transfer summary before confirming. The source must stop before the target begins; failures remain visible for recovery.
How to get started.
From your first choice to the result. Here is how to use this feature in your own project.
- 01
Review the current task’s goal, progress, changes and outstanding issues.
- 02
Choose a suitable destination route using your own available AI access.
- 03
Confirm the handover. The next agent receives the package and checks the current project state.
How it could work for you.
Hand a paused implementation to another configured engine while keeping the same project and task context.
How to assess the result
The target needs valid access. A summary cannot reproduce hidden internal reasoning from the previous model.
Where to find it.
Conversation → Tools → Cross-Agent Handover
Available actions depend on your plan, mode and project state. Features that require a task become available within an appropriate conversation.What is saved?
The handover report, destination route and follow-up task are stored.
What you need to know
Provider secrets and hidden reasoning are not transferred. External effects are not undone, and an uncertain state may need a recovery check first.