Prompts.

A prompt becomes easier to reuse when its context travels with it. This collection focuses on recording the task, expected inputs, output requirements, examples, versions, and tests that surround an instruction. It is a practical starting point for building your own prompt workflow.

Read the versioning guide to move from a saved piece of text to a record another person can understand and review. Then connect it with the AI inventory and LLM evaluation guides to document the model configuration and evidence behind a result. The examples illustrate a process you can adapt to your own materials; they are not a catalog of ready-made prompts or a promise of identical output on every run.