AI Prompt library for technical writers
Unlike the other samples in this portfolio, this page documents real tooling from my active workflow. The prompt library lives in this repository under .github/prompts/ and is fully functional with VS Code and GitHub Copilot.
As AI tools have become a standard part of technical writing workflows, I've invested in building reusable, structured prompts that produce consistent, high-quality documentation. This prompt library is designed for technical writers working in VS Code with GitHub Copilot — but the prompts are plain Markdown and can be adapted for Claude, ChatGPT, or Cursor as well.
Why a prompt library?
Ad-hoc AI prompting produces inconsistent results. A well-designed prompt library solves three problems:
- Consistency: every writer on the team starts from the same baseline, producing docs with a consistent structure and tone.
- Speed: a good prompt can generate a first draft in seconds that would take 30–60 minutes to scaffold from scratch.
- Quality control: prompts encode your style guide, preferred terminology, and structural requirements so the model produces output that matches your standards.
How to use this library
Prerequisites
- Visual Studio Code
- GitHub Copilot enabled in VS Code
- Reusable prompts enabled in your workspace settings
Setup
- Clone this repository to your local machine.
- In VS Code, go to File > Preferences > Settings and search for
chat.promptFiles. - Enable the setting and add the path to your local clone of this repository.
- You can now invoke prompts from the Copilot chat panel using
/prompt-name.
Prompts
FAQ generator (create-faq)
This prompt accepts a file of raw customer support tickets and generates a structured, publication-ready FAQ. It's designed to turn messy support data into self-service documentation without manual summarization or editing from scratch.
The prompt
---
description: Create FAQ entries from customer support tickets.
model: GPT-4.1
---
You are a technical writer on the customer support team for an e-commerce platform.
Your task is to analyze the tickets in ${input:tickets-data}, a markdown file containing
raw customer support tickets, and do the following:
- Identify the most common issues and those that can be addressed with clear documentation.
- Draft an FAQ page based solely on these issues.
- Group similar issues together to highlight patterns.
- Write clear, concise questions and answers to help customers resolve their issues independently.
- DO NOT include or infer any information that was not provided in the tickets.
- Format and organize the FAQ in markdown, using logical headings and bullet points
for clarity (e.g., Shipping and orders, Product issues).
- Name the output file `faqs.md`.
- Ensure the FAQ is user-friendly and easy to navigate.
Input variable
| Variable | Description |
|---|---|
tickets-data | Path to a .md file containing raw customer support tickets |
Example usage
/create-faq tickets-data=docs/prompt-samples/support-tickets.md
Try It Yourself
Use the test data at support tickets to test the prompt yourself. The output will be a structured FAQ in Markdown format, saved as faqs.md. Open the file to review the generated content and verify its accuracy.
More Prompts in This Library
| Prompt | Description |
|---|---|
create-faq | Generate a structured FAQ from raw support ticket data |