Verified against ChatGPT · 2026-08-09
Build a meeting agenda that forces a decision instead of producing another discussion
Converts a loose list of topics into an agenda where every item names the decision it must end in, who owns it, and what happens if the meeting runs out of time.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are building the agenda for a meeting I'm running. My real problem with agendas isn't listing topics — it's that meetings run long and end without decisions because the agenda never forced one in the first place. MEETING PURPOSE AND ATTENDEES Weekly ops sync, 6 attendees: me, two team leads, finance, and two ICs. RAW TOPIC LIST Q3 marketing budget, new vendor contract, office move update, hiring plan for support team, someone's PTO conflict. TOTAL TIME AVAILABLE 45 minutes DECISIONS THAT ABSOLUTELY MUST GET MADE THIS MEETING The vendor contract decision, since the vendor's offer expires Friday. For every topic, rewrite it as a decision statement, not a subject label — "marketing budget" becomes "decide whether Q3 marketing budget increases by 20% or stays flat." If a topic in the raw list can't be phrased as a decision because it's genuinely just informational, mark it as an FYI item and cap it at two minutes with no discussion slot — FYI items are read, not debated. Assign each decision item an explicit owner (the person who will actually say yes/no) and a time box in minutes, with the must-decide items getting priority time allocation even if it means cutting or shortening lower-priority items to fit the total time budget. Order items so the must-decide items come first, not last, since agendas that save the important decision for the end are exactly what let it get punted when time runs out. For every item, write one pre-read question the owner should think about before the meeting starts, so the meeting isn't the first moment they're confronting the trade-off. Add an explicit "if we run out of time" rule at the bottom: which lower-priority items get bumped to a follow-up versus decided asynchronously by the owner without a meeting at all. WHAT NOT TO DO Do not add generic agenda boilerplate like "introductions" or "any other business" unless I explicitly listed them. Do not pad the agenda with more items than the time budget can actually hold — cut instead. OUTPUT FORMAT A numbered agenda: item name (as a decision statement or FYI), owner, minutes, pre-read question. Followed by a short "if we run out of time" section.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Most agenda generators produce a list of topic labels because that's the shape of the raw input they're given, and GPT-5.1 will happily mirror the input's shape back unless explicitly told to transform it — asking it to rewrite every topic as a decision statement forces a structural change that a topic label can't hide behind: you can't write "decide whether Q3 marketing budget increases by 20%" without someone having to actually pick an answer, whereas "marketing budget" as an item can be discussed indefinitely without resolution. Ordering must-decide items first rather than last directly counters a known meeting-dynamics failure — items placed at the end of an agenda are the first casualties when earlier items run over, so a model that defaults to listing items in the order given (which usually mirrors whatever order they occurred to someone, not priority) will silently reproduce that failure unless explicitly told to reorder by priority instead of input order. The pre-read question per item exists because a model asked to just "add discussion notes" tends to write summary bullets restating the topic, which don't change anyone's preparation; a question specifically aimed at the owner's trade-off forces the model to identify what the actual tension in the decision is, which is a genuinely harder inferential step than restating the topic and produces a materially more useful artifact. The explicit "if we run out of time" section matters because without it, a time-boxed agenda that inevitably runs over defaults back to whichever items happen to be last, re-creating the exact problem being solved — naming in advance which items get bumped versus decided asynchronously turns an implicit, ad-hoc triage decision made under time pressure into an explicit one made calmly beforehand.
What you get back
1. DECIDE: Sign or decline the vendor contract before Friday's expiration — Owner: Finance — 15 min — Pre-read: what's the real cost of a 2-week extension request instead of deciding today? 2. DECIDE: Approve or hold Q3 marketing budget at 20% increase — Owner: Marketing lead — 12 min — Pre-read: what specific outcome would justify the increase?
If we run out of time: office move update becomes an async Slack post; PTO conflict is resolved by the two people directly involved without a meeting.
Verified against
ChatGPT GPT-5.1 · 2026-08-09
Changelog
- 2026-08-09 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — custom software, built and maintained for you — that's Scult's day job.
EXPLORE CUSTOM SOFTWARE
