I built this prompt library because I kept rewriting the same instructions for analysis, planning, coding, and communication tasks.
The library collects practical starting points for common work with ChatGPT, Claude, and Gemini. Each prompt gives the model a role, a clear outcome, useful context, and an expected format. Search and filters make it easier to find one without reading a long catalogue.
I treat a saved prompt as a template, not a magic command. Before running it, I replace generic context with the real constraints of the task. Afterwards, I check the result against source material and my own judgement.
That matters most for technical, legal, financial, or business decisions. A well-written prompt can improve consistency, but it cannot make missing context true or remove the need for review.
The hosted experiment is currently offline while I revisit how the prompt data is stored and maintained.
Legal Stuff
