LLM Prompting Library

Introduction

This repository contains a collection of prompts, templates, and resources for working with Large Language Models.

These have been developed and accumulated over the course of about 18 months of using primarily Gemini and OpenAI models. The OpenAI model usage has been mostly via a Microsoft Azure AI integration at my day job. The Gemini model usage is “everything else”.

Structure

• prompts: Organized by task.
• templates: Reusable prompt templates.
• examples: Prompt/response pairs.
• resources: Articles and papers.
• scripts: Automation scripts.
• experiments: Prompt effectiveness tests.

Naming Conventions

Files in `prompts/` are named by task (e.g., `summarization.org`). Within each file, headings are used for specific prompt variations.

Metadata and Prompt Structure (Example within `prompts/summarization.org`):

Extractive Summarization

Short Summaries

Prompt 1: Basic Extractive

The Prompt Text Goes Here…

Example Input

Expected Output

Abstractive Summarization

Bullet Point Summaries

Prompt 1: Abstractive Bullet Points

The Prompt Text Goes Here…