When you find it wrong

You will find errors. That is the point. What you do next is the difference between a near miss and a habit.

Fix the output, obviously. Then note what kind of error it was: a specific, a source, a premise, a reasoning gap. Over a month, you will see your own pattern: which kinds of task, on which tool, produce which errors. That pattern is worth more than any general advice, because it is about your work.

And tell someone. Not to complain about the tool, but because a colleague using the same tool for the same kind of work is about to make the same mistake. A shared list of "things it got wrong" is one of the most useful documents a team can have, and it takes one line per entry.

In practice. A finance team keeps a shared page called "Caught." Each entry: the task, the tool, the error, who caught it. After three months it had thirty entries and a clear pattern: two kinds of calculation the tool reliably fumbled. Those now go through a spreadsheet instead. Nobody has to remember why; the page does.

At your desk. Start your own "caught" list, personal or shared. Add every error you find with the checks, one line each. After a month, read it and write down your pattern.

Write it down.

• My first "caught" entry:
• Where the list lives:
• Who I will share it with: