Noob to Ninja Prompt Guide – HackTheSim (Free)
Beginner’s Guide to Writing Prompts
From total noob to confident creator. Learn the what, why, who, when and how. Copy the recipes, tweak them, ship better outputs.
1. TL;DR Recipe
ROLE: You are a [job/expert/persona].
GOAL: I need to achieve [clear outcome].
AUDIENCE / USER: The output is for [who will read/use it].
CONTEXT: Here is the background info: [facts, links, data].
OUTPUT FORMAT: Give me [type of output] with [length, structure, headings, bullets, code blocks].
STYLE & TONE: Use a tone that is [friendly/professional/playful/technical] and [British English, no emojis, etc].
CONSTRAINTS: Do NOT include [things to avoid]. Must include [must-haves].
EXAMPLES: Here is an example of what “good” looks like: [paste example or link].
REVISION LOOP: Ask me 3 clarifying questions before you start. After the first draft, show me 3 improvement options.
GOAL: I need to achieve [clear outcome].
AUDIENCE / USER: The output is for [who will read/use it].
CONTEXT: Here is the background info: [facts, links, data].
OUTPUT FORMAT: Give me [type of output] with [length, structure, headings, bullets, code blocks].
STYLE & TONE: Use a tone that is [friendly/professional/playful/technical] and [British English, no emojis, etc].
CONSTRAINTS: Do NOT include [things to avoid]. Must include [must-haves].
EXAMPLES: Here is an example of what “good” looks like: [paste example or link].
REVISION LOOP: Ask me 3 clarifying questions before you start. After the first draft, show me 3 improvement options.
Copy that, fill the brackets, send it. Refine after the first draft.
2. What, Why, Who, When, Where, How
- What do you want?
- Why do you need it? (Problem you are solving)
- Who is it for? (Audience drives tone and depth)
- When is it used? (Publish now, part of a workflow, evergreen)
- Where will it live? (Website, email, codebase, TikTok caption)
- How should it look? (Format, length, sections, examples, checklist)
3. The CORE Framework
Context • Outcome • Rules • Examples
CONTEXT: [who/what/where/links]
OUTCOME: I want [single clear result]. Success looks like: [criteria]
RULES: Use [style, tone]. Output as [format]. Don’t include [x].
EXAMPLES: Good: [sample]. Bad: [sample].
OUTCOME: I want [single clear result]. Success looks like: [criteria]
RULES: Use [style, tone]. Output as [format]. Don’t include [x].
EXAMPLES: Good: [sample]. Bad: [sample].
4. From Noob Prompt to Pro Prompt
| Level | Weak Prompt | Improved Prompt |
|---|---|---|
| 0 | “Write a blog post about travel” | You are a travel copywriter. Write a 900 word blog post for parents with toddlers about visiting Angkor Wat. Include 5 subheadings, a packing checklist, costs in GBP, and a short FAQ. Tone: friendly British English, practical, no fluff. End with a call to action to download our free packing list. |
| 1 | “Fix this code” | You are a senior Python dev. Analyse the following script and fix the error that makes it crash on empty input. Explain the change briefly in comments. Return only the corrected code block. Code: [paste]. |
| 2 | “Make a logo” | Act as a creative director. Generate 5 logo concept descriptions for a tech blog called HackTheSim. Each concept must include colour palette hex codes, font suggestions, and a one line brand story. Keep each concept under 80 words. |
5. Prompt Anatomy Cheatsheet
- Role/Persona: “You are a veteran UI designer…”
- Task/Goal: “Create a wireframe outline for…”
- Audience: “This is for non technical founders…”
- Inputs/Context: “Here are my notes and user quotes…”
- Output Form: “Give me a table with columns A-F…”
- Tone/Style: “Use plain English, short sentences…”
- Constraints: “No American spelling, no emojis…”
- Examples: “Model it on this sample…”
- Revision Strategy: “Ask me 3 questions first, then…”
6. Ask Better Questions
Bad: “Explain quantum computing”
Good: “You are a physics teacher explaining quantum computing to 10 year olds. Use simple analogies, one paragraph per idea, and end each paragraph with a question to check understanding. Give me 4 paragraphs.”
Not sure what constraints to add? Ask the model: “List the five most important constraints I should add to get a perfect result.”
7. Iteration Loop
- First draft – accept it is rough.
- Critique – ask the model to critique its own output against your success criteria.
- Tighten – add missing constraints, supply examples, clarify audience.
- Test – get 3 alternative versions, pick the best bits, merge.
- Lock it in – save the winning prompt in your library.
Examples:
“Improve sections 2 and 4 with real metrics and sources.”
“Shorten this to under 120 words without losing key points.”
“Suggest three hooks for the intro and let me choose.”
“Improve sections 2 and 4 with real metrics and sources.”
“Shorten this to under 120 words without losing key points.”
“Suggest three hooks for the intro and let me choose.”
8. Image Prompts (MJ, DALL·E, SD)
Structure:
Subject + Style + Medium + Colour palette + Lighting + Camera/Lens + Composition + Extra details + Negative tags
Example:
Retro futuristic city skyline at dusk, ultra wide shot, neon cyan and magenta, cinematic lighting, shot on 35mm, isometric perspective, clean vector lines, no text, no watermark
9. Code & Automation Prompts
- Provide the current code and the error or goal.
- Specify language versions, libraries, runtime.
- Ask for only changed functions or full file with comments.
- Request unit tests or sample inputs to verify.
You are a senior [language] engineer.
GOAL: [bug to fix / feature to add].
CODE: [paste].
RULES: Explain changes in comments. Return only code between “` markers.
TEST: Provide a minimal test case to prove it works.
GOAL: [bug to fix / feature to add].
CODE: [paste].
RULES: Explain changes in comments. Return only code between “` markers.
TEST: Provide a minimal test case to prove it works.
10. Troubleshooting Checklist
- Did I actually give it the facts?
- Did I define audience and tone?
- Did I show an example?
- Did I ask for a specific format?
- Did I set a length or structure?
- Did I remove ambiguity?
Ask: “Tell me what information you still need from me to do this perfectly.”
11. Prompt Library Hygiene
- Save every keeper prompt with metadata (model, use case, tags).
- Version prompts when you tweak them.
- Tag by category, audience, industry, output type.
- Store and share your best prompts on this site, or just keep them on your laptop (try Submit Prompt to HackTheSim).
12. Reusable Mini Templates
A. Rewrite & Improve
Rewrite the following [type of text] to be [tone], for [audience]. Keep it under [length]. Add [must haves]. Remove [things]. Text: “[paste]”
B. Summarise with Structure
Summarise this article for [audience] in:
– 3 bullet key takeaways
– A 2 sentence plain English summary
– One quotable line in quotes
Text: [paste]
C. Brainstorm Ideas
Generate 20 ideas for [goal] aimed at [audience]. Provide:
– Idea title (max 6 words)
– 1 sentence description
– Effort score 1–5
Return as a table.
D. Turn Chat into Action Steps
From the conversation below, extract:
– A numbered action list with owners and deadlines
– Open questions
Conversation: [paste transcript]
13. FAQs
Do I always need a role? Not always, but it helps when tone or expertise matters.
How long should a prompt be? Long enough to remove ambiguity, short enough you will reuse it. Use headings or bullets.
Why give examples? Models match patterns. Show a pattern, get that pattern.
Which model? Use the best you have for complex reasoning, cheaper ones for bulk. Test both.
