Module 3 Lesson 1: Principles of Good Prompts
The golden rules of prompt engineering: Clarity, Context, and Constraints (The 3 Cs).
Principles of Good Prompts
Prompting is an art, but it's also a science. Most "bad" AI responses are the result of "bad" prompts. To avoid this, follow the 3 Cs Framework.
1. The 3 Cs Framework
Clarity
Be direct. Avoid ambiguous language. If you want a list, ask for a list. If you want a summary, specify the length.
- Bad: "Tell me about space."
- Better: "Explain the lifecycle of a star in three paragraphs."
Context
Give the AI the background information it needs to succeed. Who are you writing for? What is the goal?
- Bad: "Write an email."
- Better: "Write a professional email to my manager, Sarah, requesting a meeting to discuss my Q4 performance review."
Constraints
Tell the AI what not to do. Common constraints include word counts, tone, and formatting.
- Example: "Do not use technical jargon," or "Provide the final answer in a JSON object."
2. The Anatomy of a Power Prompt
A perfect prompt often looks like this:
- Instruction: What to do.
- Context: Background info.
- Input Data: The text to process.
- Output Indicator: What the result should look like.
graph TD
Prompt[High Quality Prompt]
Prompt --> Clarity[Clarity: Precise Verbs]
Prompt --> Context[Context: Audience & Role]
Prompt --> Constraints[Constraints: Format & Length]
3. Use Delimiters
Use special characters like ###, """, or --- to separate your instructions from the content you want the AI to process. This prevents "Prompt Leakage"—where the AI gets confused between the task and the data.
Hands-on: Fix the Prompt
Take this bad prompt: "Help me write a blog post about dogs."
Your Task: Rewrite it using the 3 Cs.
- Role: Expert veterinarian.
- Constraint: Under 300 words, use bullet points.
- Context: For new puppy owners who are worried about diet.
Key Takeaways
- Be Direct (Clarity).
- Provide Background (Context).
- Set Boundaries (Constraints).