Assignment 1 — AI Prompt Engineering Fundamentals
Assignment 1
Open notepad or obsidian make a custom prompt template where it has
- persona (example- act like an AI expert)
- Context (example - I am a student passed out of school learning AI )
- references (give questions one by one here)
- Output format (optional) (i dont want emojis in my response , response should be obsidian markdown format, detailed simple language)
customize your prompt to get answers to these questions
- what is AI
- Rule based system vs AI
- Prompt engineering vs context engineering
- What is Agentic AI
- Difference between AI like Gemini, chatgpt, vs agentic ai
- What is LLM
- What do u mean by the line "LLMs predicts next words/token" and how LLMs does it
- what is a token
- how is AI, Machine Learning and LLM related to each other
- What is prompting ? How do we define a prompt as ac
good engineered prompt - What is zeroshot, few shot, chain of thought(COT) ,Meta Prompting
Act as a Meta-Prompt Engineer. Your task is to transform the [USER INPUT] provided below into a highly optimized, structured prompt for a Large Language Model.
### STRATEGY RULES:
1. PERSONA: Assign a world-class expert persona relevant to the topic.
2. CONTEXT: Define why this task is being performed.
3. CLEAR STEPS: Use "Chain of Thought" by providing a step-by-step execution plan (Step 1, Step 2, etc.).
4. CONSTRAINTS: Explicitly state what the AI should NOT do.
5. FORMATTING: Specify that the output must be in clean Markdown.
### OUTPUT STRUCTURE TO FOLLOW:
- **Role**: [The assigned persona]
- **Objective**: [The primary goal]
- **Context/Background**: [Detailed setting for the AI]
- **Specific Instructions**: [Numbered list of steps]
- **Constraints/Guardrails**: [Negative constraints and style guides]
- **Desired Output Format**: [How the final result should look]
### USER INPUT:
"${userText}"
---
Construct the final prompt now. Only return the final prompt text, no introductory chatter.