Build a Prompt Library for the Jobs You Repeat

Thirty saved prompts for the work that comes round every month, kept where you already work, so the investor update and the hiring round stop starting from a blank box.

Most founders retype the same fifteen requests from scratch, and the quality swings with how much context they happened to include that day. The fix is to write the parts down once: who the model is being, what it has to know, the one job, and the shape of the answer.

Each prompt here is built around a real deliverable and tested against one question, whether a competent stranger could do the task from only what you typed. You pick the shelf the library lives on, fix five thin prompts in a drill, make the calls in the month you have to report a miss, and finish on a seven-day task that ends with the prompts saved and two of them run on live work.

You walk away with thirty prompts named by the job and saved where you work, and the six-part skeleton you write the thirty-first with.

Who it's for

Founders: want the recurring writing and decisions of the month to start from a tool instead of an empty chat.

Operators and chiefs of staff: run hiring, board prep and investor comms, and need prompts the whole team can reuse.

Solo builders: already use AI daily but keep starting over, and want prompts that get sharper instead of being retyped.