Streamline Prompting with a Reusable Markdown Method

This tutorial shows how to extract any completed AI workflow into structured markdown instructions. These instructions make your tasks repeatable and transferable across different AI tools.

Who This Helps

• Marketers producing large volumes of customized content
• Entrepreneurs using multiple AI platforms
• Anyone testing and refining prompting strategies

What You’ll Create

Reusable markdown-based instructions that replicate successful AI outputs.

Getting Started

• A Google account for NotebookLM
• A sample task (e.g., generating email template variations)

Step 1: Complete Your Task in NotebookLM

Create a new notebook and perform your task as usual.

For this example, analyze existing email copy to generate custom email sequence variations.

Upload relevant documents and guide the AI through producing outputs. Be sure to define:

• Tone
• Target audience
• Word or character limits
• Brand name
• Product or value proposition
• Specific writing rules (e.g., avoid em-dashes)

Step 2: Extract Markdown Instructions

Paste the following prompt into NotebookLM:

“Take the task we’ve completed and extract the relevant context into system instructions in markdown format. The goal is to reuse these instructions across any AI with different inputs and get consistent results. Include rules such as the number of variations and any follow-up questions the AI should ask before starting.”

Step 3: Reuse Instructions Across Tools

Create a new project in NotebookLM and paste the markdown instructions as text.

You can also use them in other tools:

• ChatGPT: New Project, then three dots, and Edit Instructions
• Claude: Projects, go to New Project, then Set Project Instructions
• Perplexity: Spaces, create New Space, and Add Instructions
• Grok: Projects and Create Project

This process turns extended prompting sessions into a reusable template that delivers consistent results with new inputs.

Going Further

Store your markdown instructions in a centralized system like Notion, Airtable, or a Google Drive folder.

Each time you complete a task, save its instructions to build a reusable library. This makes it easier to delegate workflows across your team, regardless of which AI tools they use.