Instruction Following & Constraints
Learn to write clear, unambiguous instructions and apply constraints to ensure the LLM adheres to specific rules and formats.
Instruction Following & Constraints is a free Prompt Engineering & LLM Optimization for Developers lesson on CoddyKit — lesson 2 of 4. You can read the complete lesson below for free — then practise it hands-on in the browser with a built-in code editor and a 24/7 AI tutor. It is part of the Prompt Engineering & LLM Optimization for Developers learning path, one of 4 lessons in the course, and your progress syncs across the web and the CoddyKit app.
Mastering LLM Instructions
Welcome to a crucial lesson in prompt engineering! Today, we'll learn how to write super clear instructions for Large Language Models (LLMs).
- Clear instructions guide the LLM to provide exact, useful responses.
- Poor instructions lead to irrelevant or unhelpful output.
- It's like giving directions: the clearer they are, the better the result!
The Power of Clear Directives
Why are clear instructions so important? LLMs are powerful but they aren't mind-readers. They rely on your prompt to understand the task.
Ambiguous language or vague requests can lead to:
- Incorrect answers: The LLM guesses your intent.
- Irrelevant content: It might generate something you didn't ask for.
- Inconsistent formatting: Outputs vary, making automation hard.
Our goal is to eliminate guesswork for the LLM.
Specificity: Your Prompt's Best Friend
When writing instructions, think about being as specific as possible. Instead of saying 'write about dogs', consider 'Write a 3-sentence summary about the Labrador Retriever breed, focusing on its temperament'.
Key elements of specific instructions:
- What to do: The main action (e.g., 'Summarize', 'Generate', 'Explain').
- What it's about: The subject matter.
- How to do it: Any specific method or approach.
From Vague to Actionable
Let's look at an example. A vague instruction can confuse the LLM, but a specific one leaves no doubt.
Vague:
Write about apples.Improved Actionable Prompt
This improved prompt clearly states the task, subject, and desired output.
Specific:
Write a short paragraph about the nutritional benefits of apples, suitable for a health blog. Focus on vitamins and fiber content.Defining Boundaries with Constraints
Beyond clear instructions, constraints are rules that limit the LLM's output. They ensure the response fits a particular structure, format, or content requirement.
Think of constraints as guardrails, keeping the LLM's creativity within useful bounds. This is crucial for integrating LLM outputs into applications.
Shaping Output: Format Constraints
One common type of constraint is specifying the output format. This is vital when you need structured data from the LLM.
Examples of format constraints:
- JSON object
- Bullet list
- Numbered list
- Markdown table
- Plain text paragraph
Always tell the LLM exactly how you want the information presented.
Example: JSON Format Constraint
Here's how you might ask for information about a fruit in JSON format. Notice the explicit instruction for the format.
Extract the name, color, and taste of 'banana'.
Return the information as a JSON object with keys: 'fruit_name', 'fruit_color', 'fruit_taste'.Controlling Content & Tone
Constraints can also dictate the content or tone of the response:
- Content: 'Do not mention historical facts.' or 'Include only benefits, no drawbacks.'
- Tone: 'Write in a friendly and encouraging tone.' or 'Use a formal, academic tone.'
These constraints help tailor the LLM's output to your specific audience and purpose.
Check Your Understanding
Consider the following prompt. Which of the options below best describes a specific instruction or constraint applied?
Lesson Summary
Great job! In this lesson, you've learned the power of clear instructions and how to apply effective constraints when prompting LLMs.
- Instructions: Be specific, unambiguous, and tell the LLM exactly what to do.
- Constraints: Use them to control output format (JSON, lists), content (positive comments only), and length (sentence/word count).
Mastering these techniques will significantly improve the quality and consistency of your LLM interactions. Keep practicing!
Frequently asked questions
Is the “Instruction Following & Constraints” lesson free?
Yes — the full text of “Instruction Following & Constraints” is free to read here on the web, and the Prompt Engineering & LLM Optimization for Developers course includes 4 lessons in total. To practise it interactively (a built-in code editor and a 24/7 AI tutor) and unlock the rest of the Prompt Engineering & LLM Optimization for Developers course, upgrade to CoddyKit PRO.
What will I learn in “Instruction Following & Constraints”?
Learn to write clear, unambiguous instructions and apply constraints to ensure the LLM adheres to specific rules and formats. You practise Prompt Engineering & LLM Optimization for Developers with hands-on code you run directly in the browser, and a 24/7 AI tutor answers your questions as you work through the lesson.
Do I need any experience to start Prompt Engineering & LLM Optimization for Developers?
No prior experience is required. Prompt Engineering & LLM Optimization for Developers on CoddyKit is structured for beginners through advanced learners; this is — lesson 2 of 4, so you can start here or from the beginning and move at your own pace.
How long does the “Instruction Following & Constraints” lesson take?
Most CoddyKit lessons take about 5–10 minutes. Each one is bite-sized and interactive, so you make steady progress and pick up exactly where you left off across the web and the app.
Can I write and run code in this Prompt Engineering & LLM Optimization for Developers lesson?
Yes. Every Prompt Engineering & LLM Optimization for Developers lesson includes a built-in code editor, so you write and run real code right in your browser and get instant AI feedback — no local setup required.
All lessons in this course
- Role-Playing & Persona Prompts
- Instruction Following & Constraints
- Iterative Prompt Refinement
- Delimiters and Structured Prompts