Sign In
Signed out 5
Why This Matters

A prompt library is a practitioner's highest-leverage asset. Every time you find a prompt that works well for a recurring task, you're building something that will save you that discovery process forever. After six months of deliberate capture, your library is better than anything you could buy — because it's calibrated to your specific work, your voice, and your real use cases.

The Concept

What Goes in a Prompt Library

A prompt library is not a collection of clever tricks you found on the internet. It's a documented record of prompts that work specifically for your recurring tasks. The criteria for inclusion: you've used it at least twice, it produces reliably good outputs, and the task recurs often enough that having the prompt saves real time.

High-value library categories:

  • Email templates (difficult conversations, status updates, new client intros, follow-ups)
  • Document analysis prompts (extract key points, identify gaps, compare documents)
  • Writing assistance (tone adjustment, length reduction, audience adaptation)
  • Research prompts (framework generation, pros/cons, scenario planning)
  • Meeting prep (agenda refinement, pre-read synthesis, objection mapping)
  • Task-specific prompts for your professional domain

The Library Entry Format

Each library entry should have four elements:

  1. Name: A descriptive name you'll recognize when you need it ("Executive summary — skeptical audience")
  2. When to use: The trigger condition that calls for this prompt
  3. The prompt: The actual text, with [PLACEHOLDERS] for the parts that change each time
  4. Notes: What to watch for, common failures, tips for this specific prompt

Building the Library Without Extra Work

The failure mode for prompt libraries: treating it as a separate project. The sustainable approach: capture as you go.

The habit is simple: when you finish an AI session that produced something really useful, spend 60 seconds adding the core prompt to your library before you close the tab. That's it. No elaborate system required.

Tools: a plain text file, Notion, Apple Notes, Obsidian — whatever you already use for notes. The format matters less than the habit.

Team Prompt Libraries

A team prompt library multiplies individual discovery across the whole team. Instead of everyone separately figuring out how to get good AI outputs for the same types of tasks, the best prompts get shared and everyone benefits.

What makes team libraries work:

  • A shared, accessible location (Notion, Confluence, Google Doc — whatever the team already uses)
  • Clear categories that match the team's actual work
  • Each entry has a context note ("who uses this and when")
  • Regular review: prompts get outdated as AI models improve
  • Attribution: knowing who built a prompt and who to ask questions of
A library entry in practice

Here's an example prompt library entry for a common task:

Name: Status update — complex project, mixed audience

When to use: Weekly project updates to a group that includes both technical and non-technical stakeholders

Prompt:

You are a project manager writing a weekly status update. The audience includes both technical team members and non-technical stakeholders. Project: [PROJECT NAME] This week: [BULLET POINTS OF WHAT HAPPENED] Issues: [ANY BLOCKERS OR CONCERNS] Next week: [PLANNED WORK] Write a status update that: - Opens with the one-sentence takeaway a non-technical executive needs to know - Covers technical detail in a separate section labeled for the team - Is direct about issues without creating alarm - Is under 300 words total

Notes: If there are no issues, still include the section with "None this week" — removing it makes it look like you forgot. The one-sentence executive summary is always the hardest part; spend time on it.

Hands-On Exercise

Start your prompt library with 5 entries

NotionObsidianApple NotesAny notes tool
Look back at the past two weeks of work. Identify 5 tasks where you used or could have used AI. For each one, write a library entry: 1. Name (descriptive, specific) 2. When to use (trigger condition) 3. The prompt with [PLACEHOLDERS] for the parts that change 4. One note about what to watch out for Put these somewhere you'll actually find them again — not a new tool, whatever you already use. Then: share one entry with a colleague and get their feedback. Does it work for their version of the same task?
Five is the right number to start. Ten feels like a project. Five feels like a useful list.
Active Recall

Before moving on — close this lesson and answer these from memory. Then come back and check. Testing yourself (not re-reading) is how this sticks.

1 What are the four elements of a useful prompt library entry? Why does each one matter?
2 What's the sustainable habit for building a prompt library without it becoming a separate project?
Reflection

What are the 3 tasks in your work you do most often that involve AI or could involve AI? These are your highest-priority library entries. Write them down now.

Key Takeaway

A prompt library built from your own recurring tasks is more valuable than any generic collection. Capture as you go — 60 seconds after a successful AI session. Four elements per entry: name, trigger, prompt with placeholders, notes. A team library multiplies individual discovery.