Modern models are good enough that one-sentence prompts often get 85% of the value with 1% of the effort. The barrier to getting started is lower than people think.
For most things, you don't need a big complicated prompt to get really good results in cowork or codex and more recent models.
A year or three ago you did. Now Opus 4.7 (or even Sonnet 4.6) and GPT 5.4 / 5.5 are really good at delivering a great result without much agonizing over the instructions and I think that's a barrier for a lot of folks getting started.
Here's my sales call coaching feedback prompt I posted about recently that runs in cowork every day:
"Check for google calendar meetings in the last day that fit the criteria of a sales call. Pull the transcript from granola and give me feedback on how well I did"
That's the whole thing. I'm sure I could tune it to give myself even better feedback, but this is more than good enough. It's not even 80/20, it's like 85/1. I have almost 10 of these little routines running daily or weekly and am experimenting with new ones all the time to see what sticks.
Get More Like This
Follow along as I build and share what I learn
Found this helpful? Share it with your network!