Switch AI roles without losing the thread of the project

Projects often need a handoff.

In short

When you switch roles, hand over the context that matters for the task so a better-fit AI can continue without making you explain the project from the beginning.

One role may be better at analysis, another at implementation. If one role gets stuck, you may want to hand the problem to someone stronger at that part. If every handoff requires retelling the project, multi-role work quickly becomes more repeated coordination.

Hand over the task, not the entire history

FinchStory is making role handoffs more natural. Roles can continue around the same project context without making you restate goals, constraints, files, and every decision that is already settled.

The value of multiple roles is easier takeover

That does not mean every role needs to see everything. The useful handoff is the context the next role needs for the task, not a blind copy of the entire conversation history.

The point of multiple AI roles is not simply having more AI. It is being able to move the project to a better-fit role when the work changes.

What changed

Keep the context that matters when switching roles.
Reduce repeated explanation of goals, constraints, and files.
Let analysis, implementation, and review roles take over in sequence.
Avoid copying the entire history to every role by default.
FinchStory 0.9.1 · September 10, 2026