Great minds

Write me a skill called munger-test, for a big decision I'm about to make. Write its description as the sentences I actually say, like "Munger test this" and "how would this fail", and say it is for a price rise, a hire, a new supplier, a programme I'm being sold or a big client, never a small choice. When it runs, it does this, in order: 1. Says my decision back to me in one sentence, and what I want to get from it, and asks me one question if either is unclear. If the decision is small and easy to undo, it says so and keeps the whole report short. 2. Reads what this folder already holds about my business before it writes anything, and tells me which files it read. 3. Turns the question round, the way Charlie Munger did, and lists the surest ways this decision fails, the likeliest first and no more than seven. It draws them from what my files show, from how decisions like this usually go wrong for other small businesses (labelled as general knowledge, never with made-up figures), and from Munger's checklist of what might be fooling me: someone who is paid if I say yes, a deadline, overoptimism, doing it because others do, liking the person selling it, carrying on because I've started, deciding while stretched, or needing a skill I haven't got. It names only the ones my files show, and says when several push the same way. 4. Checks my plan against each one and tells me whether my plan already does it, might do it or avoids it, or that it's unknown if my files don't say, with the line from my files that shows it. 5. Turns each way to fail into a rule for what not to do, with the cheapest fix or check before I commit and who can answer it. 6. Ends on the two my plan is closest to, says plainly what my plan already avoids, and leaves the decision to me. It keeps the report under about 600 words, in short lists rather than tables, quotes my files and labels anything that is an estimate, and it writes nothing else and sends nothing. Show me the file.

I'm about to make a big decision. Test it the way Charlie Munger would, by turning it round. Ask me what the decision is and what I want from it, one question at a time. Then list the surest ways to make it fail for a business like mine, likeliest first, including anything that might be fooling me, like someone paid if I say yes, a deadline or overoptimism. Check my plan against each, turn each into a rule for what not to do, say what my plan already gets right, and end on the two it is closest to. Never tell me whether to go ahead.