The printable CI toolkit.

Map your life once, on one page: where the work comes from, what it is, how you handle it, what good looks like, and who it's for. Fill left to right. Twenty minutes, not an afternoon.

Check: does every output serve a named customer? If not, drop it or park it.

Once you know your customers, map what each one needs from you. Start with the C column of your SIPOC. Ideas: family, spouse, work team, close friends, guests, and yourself.

For each customer from your SIPOC, write what they would actually say, the need underneath, and a plain, checkable standard for “good” (critical to quality).

A good CTQ is something you could check on a calendar or a count. “Be more present” → “dinners without screens, most nights.”

Every small thing points up. Write top to bottom; each level should name the one above it serves.

Files, inboxes and notes, organized like a well-run workshop. Do the first pass once; then keep the weekly shine.

Decide exactly what counts before you measure. Then measure two ordinary weeks, before you change anything.

Write the definition before you look at the number.

Walk one real loop with real data (screenshots, a week of inbox and calendar). For each step, write what really happens and tag the waste with a DOWNTIME letter.

Tally a month of one thing (emails by kind, tasks by type, hours by lane). Sort biggest first, add a running total, and circle the few that make about 80%.

When something slips, ask why until you reach a cause you can fix. Stop when the answer is a process, not a person.

Write the problem in the middle. Brainstorm possible causes under each category, then circle the likeliest and test it with 5 Whys.

The current best way to run a routine, written down. If a step has a clear “done when,” a routine can eventually run it for you.

One card per item. Set a small limit on Doing and on Waiting on you; when a column is full, nothing new moves in until something moves out.

Spread the week evenly: a little deep work, admin, people time and home each day, instead of spikes. Keep one day free.

Any day marked heavy? Move one block to a light day before the week starts.

One row per recurring task. Exactly one A per row. If it leaves the house, the A is you.

Put the one-off in the middle. Fill the branches before you hand it to a deep-work partner; the branches become the brief's headings.

The A3: the same brief on one page. The mind map's branches fill it in.

For each agent or routine: what could go wrong, how bad, how likely, and the guardrail that prevents it. Re-check when a new bot joins the team.

Standard guardrails: drafts not sends (poka-yoke) · stops and asks (jidoka) · heartbeat, fail loud.

For each Dashboard measure: what normal looks like, who checks it, the signal that means act, and the reaction.

Signal vs. noise: one odd week is noise. A run of three drifting the same way, or a point outside the range, is a signal.

One glance, not a read. Sort this week's items by color, put anything red in “Waiting on you,” and keep the page short.

One page that tells you whether the system is working. Pick three to six measures, tie each to your future self or a macro-plan theme, and review them weekly in the week-ahead, not daily.

Monthly or quarterly. List what you say matters (from your future-self values), then pull the facts from your calendar or a time audit. No grades: one observation, one small experiment, check again next month.

Every time you edit or turn down a draft, jot what you changed and what the system should keep. Then save it where your AI will read it next time.

Examples: “warmer, shorter” → style · “never say X” → rule · “always cc Y” → preference.