Do not calculate this for me. Build me a calculator I can change the inputs on.\n\nThe decision: [DESCRIBE IT -- which pricing plan, whether to hire, rent versus buy].\n\nThe calculator needs:\n\n- An input for every variable that genuinely moves the answer, and none for the ones that do not\n- Sensible starting values, with a note on where each one came from\n- The break-even point shown explicitly, not buried in the output\n- A clear statement of which input the answer is most sensitive to\n\nAfter you build it, tell me in one line what the answer is at the default values, and what would have to change for the other option to win.
Here is data copied out of [WHERE -- a spreadsheet, a report, a dashboard]. It is messy.\n\nBefore charting anything, show me what you had to fix: inconsistent units, missing rows, duplicated entries, dates in more than one format, anything you inferred. List it so I can check you inferred correctly.\n\nThen build the chart that answers [THE QUESTION], and make it editable so I can change the grouping.\n\nPick the chart type for the question, not for looking impressive. If the honest answer is that this data does not support the question I am asking, say that first and tell me what data would.\n\n[PASTE DATA]
Build me an interactive model for [SITUATION -- runway, hiring plan, pricing change, capacity].\n\nThree scenarios: pessimistic, expected, optimistic. Let me adjust the assumptions behind each one rather than just picking between them.\n\nShow:\n\n- The assumptions as editable inputs, each labelled with how confident it is reasonable to be about it\n- The outcome for all three side by side\n- The single assumption that changes the answer most, called out explicitly\n- The point at which the pessimistic case becomes a problem I have to act on\n\nDo not smooth the pessimistic case to be reassuring. I need the version I would plan against.\n\n[PASTE WHAT YOU KNOW]
I need to collect [WHAT] from [WHO]. Build the form, do not describe it.\n\nRequirements:\n\n- Only fields that change what happens next. Challenge any field I asked for that fails that test.\n- Validation on each field, with the error message the person actually sees\n- The order that gets the highest completion rate, with your reasoning\n- Which fields are optional, and what it costs me when they are left blank\n\nThen tell me the one question most likely to make someone abandon this form, and whether it is worth asking anyway.\n\n[DESCRIBE WHAT YOU NEED TO COLLECT]
Turn the plan below into a visual timeline.\n\nI care more about the dependencies than the dates. Show:\n\n- What cannot start until something else finishes, drawn as a link not a note\n- The critical path, marked clearly\n- Every point where one person or team is needed in two places at once\n- The slack in each non-critical task\n\nMake the dates adjustable so I can see what moving one thing does to everything else.\n\nIf the plan as written is not achievable in the stated time, show me that rather than compressing tasks to make it fit.\n\n[PASTE PLAN]
I keep checking [WHAT] to answer [WHICH QUESTION]. Build me a single view that answers it at a glance.\n\nRules:\n\n- One number is the headline. Pick it and defend the choice.\n- Everything else is there to explain that number or to show when it is misleading\n- Include the comparison that makes the headline meaningful: versus last period, versus target, versus a benchmark\n- No metric that I would not take an action on. If it is interesting but not actionable, leave it out and tell me why\n\nTell me what this view would NOT show me, so I know when to look somewhere else.\n\n[PASTE DATA OR DESCRIBE THE SOURCE]