Who this is for

You’ve run the basics and want the full surface — every door, when each one earns its place, and the traps.

The seven doors

Each is a typed command. You can also just say what you want in plain language — Claude picks the same playbook either way.

DoorThe momentWhat actually happens
/flow:setupfirst run, or something’s brokenChecks the connection, whether your horizons are written, and whether your foundations are there. Only the first two can block you
/flow:triage”I’m drowning”Disposal. Every task gets exactly one home. Works on day one with nothing but tasks
/flow:plan-my-day”what should I work on”Selection against the week’s story, inside honest capacity
/flow:focus”I can’t make myself start”The declaration, then the timer
/flow:shutdownend of the dayNothing unplanned left with a hook in it, and tomorrow’s first move set up
/flow:wrapend of weekAccount → gap → homes → direction
/flow:loop-check”have I dropped something?”Completed work whose outcome never landed

The rhythm they’re built for

Daily — plan, then focus, repeatedly, then close. The plan hands off into a declaration rather than a bare timer, so the day turns into work rather than a list; the shutdown makes sure nothing you’re still carrying follows you into the evening unplanned.

Weekly — the WRAP. It closes the week and writes the next one, carrying forward whatever the gap taught you.

When needed — triage when the pile gets loud, loop-check every few weeks.

Rarely, and the plugin decides when — re-meeting a stale horizon, and noticing a person who keeps appearing in your reflections. These ride inside the WRAP and fire on a trigger. If they showed up weekly they’d be meaningless, so they don’t.

It can start your real rituals

BeginWell, Shutdown, and the WRAP live in MakeTimeFlow, and they’re fuller than anything a conversation can run — the Shutdown alone walks your waiting-fors, the messages that need a last look, and an actual stop.

So the plugin doesn’t just recommend them. It can start one: it asks how long you want to give it, then opens the ritual in the app with a timer already running.

That question is worth more than it looks. Almost everyone underestimates a shutdown, and naming a length up front is what stops a ritual quietly becoming a five-minute skim. Fifteen minutes for a Shutdown, ten for BeginWell, thirty or more for a WRAP are honest numbers.

It opens in the app

The ritual opens in MakeTimeFlow, not in the conversation. If the app isn’t open, what you get is a timed task that opens the ritual whenever you start it — still useful, just not immediate.

If you’d rather stay in the conversation, /flow:shutdown and /flow:wrap run their own shorter versions. What those write is saved to your journal alongside your ritual — they don’t mark it complete, and the plugin won’t claim they did.

The part that makes it different

Every other planning tool works from your task list. This one works from your foundations as well, and that changes the answers:

Your values break ties, out loud. When two tasks both fit the day and both serve the week, the plugin picks using your own named values, in your own order — and tells you which one. Repeatedly meeting your own values attached to real choices is the mechanism; a values list nobody revisits is decoration.

Your horizons get maintained, not just set. Everyone helps you write a five-year vision. Almost nothing helps you keep one alive — and a vision nobody has read in eight months has quietly stopped guiding anything while still looking filled in. When one goes stale, the plugin reads it back and asks whether it’s still true. It doesn’t rewrite it.

Your strengths route work; they never label you. On a day when nothing appeals, the read is mis-assignment, not motivation: none of today’s tasks fit how you actually work. That’s a scheduling observation, not a personality claim, and the plugin is built not to make the second kind.

People are noticed, never tallied. The evenings that matter most are exactly the ones no tracker sees, so any ledger of relationship time understates the relationships that matter most. Instead it occasionally asks about someone who recurs in your own writing.

If your foundations are empty

The plugin never gates your day on them and never runs a setup wizard. Instead it offers to mirror and correct: it reads what you’ve already written and done, proposes two or three values it thinks it sees, and lets you fix them. Correcting a wrong guess takes five seconds; composing five values into a chat window is the tax nobody pays.

Offered once, easy to decline, and there’s a picker if you’d rather not do it in conversation.

How writes work

Reads are free; every write is drafted first. You see the tasks, the blocks, or the story before anything is saved.

Triage and the WRAP draft the whole set and take a single yes — confirming forty dispositions one at a time turns five minutes of relief into an interrogation.

Two rules that never bend:

  • Success stories are whole-document writes. Claude reads the current version, edits within it, and sends it all back — so it can’t append a second copy or clobber an edit you made in the app.
  • Committing a story is your gesture alone. Writing one never activates it.

Traps worth knowing

  • “Due this week” and “done this week” are different questions. The WRAP reads what you finished; a filter on what was scheduled would report a different week.
  • A follow-up completes the original task. It means “that’s done, here’s what’s next” — not “park this for later.” Triage uses buckets for parking.
  • Cancelling is recoverable. It’s a soft delete, so “not happening” is a safe answer rather than a destructive one.
  • Waiting-for has no since-date. It’s a tag, not a field, so the plugin asks how long it’s been rather than pretending to know.
  • A quiet follow-up chain isn’t proof you dropped something. Loop-check surfaces candidates; most completed work is simply complete, and loops closed outside the system leave no trace.

What isn’t here yet

Named honestly, because knowing the edges is part of using it well:

  • No Flourishing Map. The plugin sees your work — tasks, time, stories, values. It can’t yet see the six dimensions, so three of them go unrepresented.
  • No Personal Experiments or Smart Bets. Bounded experiments are core to the method and the plugin can’t run one yet.
  • No quarterly design. Horizons get maintained; authoring a quarter well is still a job for the app.