Tasks
Plans and delegation
How longer work gets organised, a plan doc with live to-dos, follow-up tasks for other staff, and work scheduled for later.
When a request is bigger than a single reply, the assistant stops improvising and organises. You can watch the whole thing happen.
Where it runs: in a task, and in AI staff runs.
What the assistant can do
| It can | What that looks like |
|---|---|
| Write a plan | A plan doc beside the task, with its to-dos listed at the bottom |
| Tick the plan’s to-dos | Items tick off in the plan doc as the work gets done |
| Delegate a task | Creates a separate follow-up task for another staff member, fire and forget |
| Create a task | A new work item in the workspace for someone to pick up |
| List tasks | Checks your open tasks, mostly to avoid creating a duplicate |
| Post into the thread | Shares progress or an intermediate result with the humans watching |
| Schedule itself | Comes back later with a given instruction, a “check back” trigger |
| Wait | Pauses briefly for something just triggered to settle |
The plan
Ask for a plan, “plan this”, “scope it”, “map it out”, and you get a plan doc beside the task: the goal, what the assistant is working from, the process phase by phase, what could make it harder, and what you end up with. It is an ordinary doc, so you can edit any part of it, and nothing starts until you say go, reply in the chat or press Run plan on the doc.
A task has one plan. Asking for changes rewrites that plan as a new version rather than starting a second one.
The to-dos
At the bottom of the plan doc is its To-dos list: one checkable step per line. It is not decoration, it is what the assistant is actually working from. While it works, the step in progress shows a spinner and finished steps tick off, live.
The list is yours too. Tick a step, click one to rename it, remove it, or add one with New. If the plan is wrong, say so immediately; correcting the plan is far cheaper than correcting the work.
Small jobs don’t get a checklist, only a plan doc has to-dos.
Delegation
A delegated task stands on its own after this conversation. It shows up in the task list assigned to another staff member, runs by itself, and returns nothing to the conversation that created it. When the result is needed to finish the current job, the assistant does that step itself.
Scheduling and waiting
A staff member can defer its own work, “check whether the supplier replied tomorrow morning”, by scheduling a later run with an instruction to itself. Short pauses (a few seconds, waiting for something it just triggered to settle) are a separate, much smaller thing; anything longer becomes a scheduled run rather than a held-open session.
Working with it well
Correct the plan, not the output. The plan is visible before most of the cost is incurred.
Say when something is a follow-up. “Do the analysis now, and have someone draft the client note afterwards” maps cleanly onto doing the work plus a delegated task.