When a plan keeps disappearing from your calendar, the first useful question is where the decision breaks down. You may be trying to describe a direction, compare two specific commitments, or keep a commitment you have already chosen. Those problems need different inputs and produce different documents. Use the following workflow with a private notebook or a redacted copy in an assistant you trust.
Map a situation before naming a goal
Take Mei, who says she needs a career change after three weeks of late support shifts. Before opening a course catalogue, she records two ordinary weeks. For each recurring commitment she notes the time, how predictable it was, whether it gave or cost energy, and who could change it. She marks a medical appointment and family pickup as fixed and leaves employer and client names out of the record. At the end of the fortnight she has separate observations about shift volatility, the work she enjoyed, and what she could actually afford to change.
Ask an assistant to turn those observations into three columns: what happened, what I inferred, and what I still need to ask. An entry such as “three shifts were moved with less than a day's notice” goes in the first column. “I need another profession” is an inference. “Can the team post a stable rota two weeks ahead?” is a question. If the assistant fills an unknown with a plausible story, remove it. The output is a one-page map with one constraint, one preference, and one testable question for each area that matters. It gives you something to show a mentor without handing over your diary.
Compare one decision at a time
Mei can either try an internal transfer or apply for a part-time course. She sets a decision date, lists deadlines and refundable deposits, and works backward to find the latest date on which each choice is still open. She also keeps a third option: ask for a four-week fixed rota and defer the more expensive commitment. For each option she writes the first reversible step, the cost of reversing after that step, and the person who can confirm the facts. A recruitment brochure is evidence about the published course schedule. It cannot establish that her manager will accept an accommodation.
Use hard gates before preferences. If two evenings are unavailable, an option requiring both evenings needs a verified schedule change before comparison. Among the remaining options, test what would flip your choice. Mei changes only one assumption at a time: first the transfer date, then weekly study time, then the amount she can lose on fees. When a small change flips the result, the output is a research task rather than a recommendation. She requests a sample timetable and a conversation with someone who moved teams. Her decision memo records the evidence, the open question, and what she plans to do if it remains open on the deadline.
Put a small experiment on the calendar
Mei chooses to test an evening of study before making a deposit. She reserves Tuesday 19:30–20:10, downloads one public practice exercise, and defines done as a short exercise plus a note on what was harder than expected. The fallback is 15 minutes on Saturday with the same exercise. A calendar event can show that time was reserved. Her completed exercise is the evidence that the test happened. She keeps both, together with a reminder to recheck the application deadline.
After two weeks, compare planned attempts, completed artifacts, and disruptions. If the only missed slot was replaced by mandatory work, test a different window. If every attempt stalled at choosing material, prepare the next exercise in advance. If she learns the work itself is unappealing, capture that observation and revise the decision memo. Do not treat a failed trial as a reason to schedule more reminders. The review should produce exactly one change to the next experiment and a date to assess it.
Choose a Skill by the output you need
Search for situation mapping when your inputs are observations and your desired output is a structured inventory. Search for a decision worksheet when you already have alternatives and need constraints, evidence and reversal costs on one page. Search for weekly planning or a habit review when you can name the commitment and need a schedule and a record of attempts. The three categories are useful even in a plain text editor. For a Skill, read its published instructions and source first. Look for the required inputs, where any record is stored, and whether it can create reminders or merely draft their text. Test a made-up example before entering personal details. A real calendar entry requires a separately connected calendar and a check of the time zone and recurrence.
Practice with a small decision
Try choosing between joining a Saturday ceramics class and practicing at home. Write the budget cap, travel time and refund deadline, then collect a real sample session time. Create one evidence row for each claim and mark whether the evidence expires. Instead of asking which option suits your personality, ask for two observations that would change the order of the choices. Carry out the cheapest test, such as completing a small project at home using borrowed tools with permission. Keep a photo or short note as the artifact. At review, decide whether the class solves a difficulty you actually experienced, such as access to a kiln or feedback on technique. Save the note next to the decision so you can explain why you changed your mind later.
Troubleshooting and FAQ
If your assistant ranks everything equally, ask it to separate commitments you cannot move from preferences you can trade. If the journal is too personal, replace identifying details with role names and use rounded time or cost ranges. If a decision cannot wait, make an explicit provisional choice with a review date and write down which facts you were unable to confirm. If the weekly review takes too long, retain only the planned attempt, the actual result and the next adjustment. You can stop the experiment once it answers the question it was meant to test.