10 patterns that 10× any LLM output

prompt-engineering-cheat-sheet · library

How it works

Pick the prompt

Fill the brackets

Run in any LLM

Get pro-grade output

Why it beats the alternative

This product

Writing prompts from scratch

Time to output

Seconds — fill the brackets & run

Hours staring at a blank box

Quality

Proven, pro-grade, repeatable

Hit-or-miss, inconsistent

Coverage

Every job already written

Reinvent each one yourself

Ownership

Reuse on every client

One-off each time

How it actually works

Exactly what it does — and when.

1For vague outputs like 'write a blog post' → it shows the before/after where adding role+goal produces a targeted draft, using the example prompt 'Act as a B2B copywriter, write a 300-word LinkedIn post about SaaS churn.'

2For controlling structure and format → it uses output-shaping to constrain replies to JSON or bullet points, using explicit delimiter instructions like 'Return ONLY valid JSON with keys: title, summary, tags.'

3For tasks where the model guesses intent → it applies one-shot prompting by embedding a single concrete example, using a labeled input/output pair separated by clear delimiters.

4For multi-step reasoning that degrades when rushed → it chains drafts by requesting an outline first, then a draft, then a critique pass, using sequential prompts on models like GPT-4 or Claude.

In practice · Fixing a rushed AI draft

A marketer pastes their existing prompt into the universal template, layers on the chain-of-drafts pattern for a product launch email, and gets a structured outline before the full draft. The before/after examples show exactly why the original prompt returned generic filler, so they know which pattern to grab next time.

Use it to…

Repair an ambiguous prompt

Use the checklist to add missing context, constraints, output shape, and acceptance criteria.

Standardize a team prompt

Turn a frequently repeated request into a shared, editable prompt pattern.

Review a model request

Check whether the prompt exposes assumptions and asks for evidence before running it.

The shift it creates

Without it

Staring at a blank prompt box
Generic, inconsistent output
Rewriting the same prompts daily

→

With it

Apply role+goal, output-shaping, one-shot, and chain-of-drafts to turn fuzzy prompts into precise, ready-to-use answers on the first try
See exactly what changes with a before/after example for every pattern
Stack patterns with the one universal template across ChatGPT, Claude, Gemini, or any model you already use

What's inside

10 high-leverage patterns

Before/after examples

Universal stack-the-patterns template

Free — no email, no payment

Who it's for

Knowledge workers drafting emails, reports, and summaries with AI

Vague replies waste editing time — these patterns shape the output before you ever hit send

Developers and builders prototyping with LLMs

One-shot and chain-of-drafts cut trial-and-error when you need reliable, structured responses

Anyone new to prompt engineering

10 named patterns with concrete before/after examples teach the craft fast — no abstract theory

Questions, answered

✓ Free · no email✓ Yours to own✓ Use on every project✓ Built to ship results