plan-day

The point is realism. A plan that doesn't fit the day is worse than no plan.

Steps

1. Run the gathering half of /daily-brief if it hasn't run this session.
2. Read brain/routines/daily.md for the energy pattern, and brain/identity/goals.md for the current single focus.
3. Find the actual free blocks between calendar commitments. Subtract the non-negotiables from daily.md. What's left is the real budget — usually much less than expected, and saying so is the job.
4. Estimate each candidate task honestly. If the tasks total more than the budget, say so explicitly and cut, don't compress.
5. Match work to energy: hard thinking into the sharp window, admin and calls into the slump.

Output

You have <X> hours of actual working time today, in <n> blocks.

09:30–11:00  <deep work>   → <task> · <why now: it's the focus / it's due>
11:00–11:30  buffer
11:30–12:00  <admin>       → <2–3 small things, batched>
...

NOT TODAY — and that's fine:
- <task> → <when instead>

ONE THING: If only one item gets done today, make it <task>.

Rules

• Leave 20% unbooked. Days that are planned to the minute fail by 10am.
• Never plan over a non-negotiable from daily.md.
• Push back if the day is unrealistic — that's the value, not the schedule itself.
• Don't create tasks or calendar blocks unless asked. Offer.