Wren

I tested this because I believed the same thing. Recorded the same vocal take with a 90 mic and a 700 mic in an untreated bedroom, then hung four moving blankets and did it again. Every listener I played it to picked the cheap mic in the treated room over the expensive one in the bare room, and it was not close. The room is printed onto the recording and you cannot remove it later. The mic difference is real, it is just smaller than the room difference until the room is fixed.

Whichever one you will still be doing in March, and honestly that is usually the one that takes under thirty seconds.

What matters is not the tool, it is recording a small number of things every day rather than a lot of things on bad days. If you only log the bad days you have no baseline and the data tells you nothing.

My five columns, filled in at the same time each evening:

• Sleep hours, roughly.
• Migraine yes or no, plus a 1 to 3 severity.
• Caffeine, count.
• Meal skipped, yes or no.
• One free text word if something unusual happened.

That lives in a spreadsheet on my phone because I can export it and sort it. Apps are lovely until you want to look at twelve months at once.

It is the difference between a diary and data. The boring days are your control group.

Yes, you're supposed to point it at the files. 40 tool calls means it spent most of that time doing the part you had already done: working out where things live.

For a task where you already know the answer: name the files, name the pattern to copy, say what done looks like. "Add a vatNumber field to CustomerForm.tsx following how taxId is done, update the zod schema in schemas/customer.ts and the handler in api/customers.ts." That's about eight tool calls and 40 seconds.

Exploration mode is for when you don't know where things are, unfamiliar repo, a bug you can't localise. Then 40 tool calls is cheap and worth it. The mistake is using exploration mode on a task you could have specified in a sentence.