NIOO Prompt Library

Table of contents

• Introduction
  • Why this library, why now
  • What a prompt is
  • Why a shared library matters
  • What you will find inside
  • How to start in five minutes
  • How to contribute
  • Principles
• 1 Basics
  • 1.1 Zero-Shot Prompting
  • 1.2 Few-Shot Prompting
  • 1.3 Role-Based Prompting
  • 1.4 Audience-Tailored Prompting
  • 1.5 Structured Output Prompting
  • 1.6 Chain-of-Thought Prompting
  • 1.7 Tree-of-Thought & Prompt Chaining
  • 1.8 Meta Prompting
• 2 Prompts for NIOO-Specific Use Cases
  • 2.1 Summarizing Ecological Literature
  • 2.2 Reviewing Grant Proposals
  • 2.3 Translating Field and Lab Notes into Structured Data
  • 2.4 Communication and Outreach Texts
  • 2.5 Policy and Stakeholder Briefings
  • 2.6 Education and Citizen Science Support

Introduction to the NIOO-KNAW AI Prompt Library

Why this library, why now

The way we work is changing. AI is becoming part of everyday research and operations, and prompts are the way we interact with it. At NIOO, we want to use these tools to keep up with change, and where possible, to improve the quality and output of our research.

This library helps us capture what works, so knowledge is not scattered but shared across the institute. Alongside the upcoming AI policy, which sets the boundaries for responsible use, this library offers practical examples and inspiration to make AI genuinely useful in our work.

---

What a prompt is

A prompt is simply an instruction in plain language that tells an AI what you want. Prompt engineering is the skill of being clear and specific. The clearer the task, the better the result.

Six building blocks make for a strong prompt:

• Task: What do you want the AI to do?
• Context: What background should it take into account?
• Audience: Who is the output for?
• Persona: From what perspective should the AI respond?
• Format: How should the answer be structured?
• Tone: What style of communication do you need?

Each of these will come clear through examples in the chapters that follow.

---

Why a shared library matters

Working together and sharing knowledge is the main thing we do. With a shared prompt library, we can:

• Accelerate knowledge across the institute by reusing proven prompts.
• Experiment with confidence knowing others have tested similar ideas.
• Blend research and operations by capturing prompts that work for both scientists and support staff.
• Expand over time: today we start with basics for tools like ChatGPT, but we welcome prompts for ecological data analysis tools, literature searches, or other software that proves useful.

If you’d like to request new additions, just email RSO@nioo.knaw.nl.

---

What you will find inside

After this introduction, each chapter explains a core prompting strategy and then gives you ready-to-use examples in code blocks. You can copy-paste these into your AI tool of choice and adapt them for your own work.

codeblock Example placeholder:

---

How to start in five minutes

1. Pick a tool. You might try ChatGPT, Gemini, Le Chat by Mistral (EU-based❤️), or another large language model.
2. Think about your application. Are you summarizing a paper, drafting an email, or designing an experiment?
3. Pause before you paste. Do not put sensitive information — names, finances, or groundbreaking unpublished research — into free tools. Remember: free doesn’t mean risk-free, and some providers may train on your data.
4. One practical ALWAYS-do tip: Go to your tool’s data/privacy settings and turn off “improve the model for everyone.”
5. Run, reflect, refine. See what works, tweak a little, and save good prompts to share back.

---

How to contribute

This library grows as we do. If you discover a useful prompt, or want to add prompts and ideas for new tools, just email us at RSO@nioo.knaw.nl or stop by. Together we’ll make sure these contributions are easy for colleagues to find and reuse.

---

Principles

The principles for responsible AI use at NIOO-KNAW will follow in the upcoming AI policy. For now, the key idea is simple: experiment, share, and stay mindful of data privacy.

---

Start

Go to chapter 1 or chapter 2