Mastering Prompt Engineering Revision
Mastering Prompt Engineering Revision
Speak to AI Like a Pro
Framework: CO-STAR · Tools: Google Gemini, OpenAI ChatGPT
The core idea
A prompt is the instruction you type into an AI tool, and the quality of what you get back is directly tied to the quality of what you ask. Most beginners write one vague word or phrase and wonder why the output disappoints — AI needs the same context you'd give a new colleague to do a task well.
The CO-STAR framework gives structure to a prompt so you don't have to guess what to include: Context (background the AI needs), Objective (the specific task), Style (the writing approach), Tone (the emotional register), Audience (who it's for), and Response Format (how it should be structured). You don't need all six every time — simple tasks need none of it, medium tasks need Objective + Tone + Format, and complex tasks benefit from the full six.
Two techniques multiply CO-STAR's effectiveness: Chain-of-Thought prompting (adding "think step-by-step" to complex requests, shown by Google Research to meaningfully improve reasoning accuracy), and iteration — refining a response with a follow-up instruction instead of deleting and restarting from scratch.
Key terms
| Term | Definition |
|---|---|
| Prompt | The instruction or message you give an AI tool to start or steer a conversation |
| Role assignment | Telling the AI to act as a specific expert or persona before answering |
| Iterative prompt | A follow-up instruction that refines a previous AI response rather than replacing it |
| Chain-of-Thought | Asking the AI to reason step-by-step through a problem instead of jumping to a conclusion |
Common mistakes to avoid
- Typing a single vague word or phrase and expecting a tailored result
- Deleting a close-but-imperfect response and starting over, instead of iterating on it
- Never specifying an audience or format, leaving the AI to guess how the output should look
- Treating every prompt as needing the full CO-STAR framework, even for simple one-line requests
Before the exam, make sure you can...
- Know all 6 CO-STAR elements and what each one contributes
- Be able to write a prompt using at least 4 CO-STAR elements from scratch
- Understand why Chain-of-Thought prompting improves accuracy on multi-step tasks
- Know the difference between restarting a conversation and iterating on it