SMARTCHAIN and LYRA
Prompt frameworks for beginners and experts
Now accepting keynotes for Q3/Q4 2025
The SuperSkills Era: Thriving in the Age of AI - I’m delivering 60-minute executive sessions for senior teams based on the research and frameworks from my upcoming book. Available across the UK, Europe, and globally in person or virtually. To book for your board, corporate team, or keynote event, fill in this form.
In addition to this newsletter, I recommend some other great ones. All free. Check them out here.
Friends,
For beta readers of SuperSkills, you should have a separate email from me. Please check your spam
I’m on holiday this week, soaking up some culture and switching off, so no reading list today.
Increasingly in speaking to groups, I see groups of individuals who usually represent a scale of AI usage.
1 Power users – employer-backed, deep into AI.
2 Tinkerers – try ChatGPT or Grok now and then.
3 Searchers – use AI like Google.
4 Starters – haven’t begun.
If you’re in group 4, open ChatGPT today and ask it to summarise a news story you just read. That first step matters.
For those in group 1, you’ll see a reference to LYRA at the bottom of this email. Skip this next section and go there.
For those in groups 2 and 3, I encourage you to try SMARTCHAIN, which is my approach to making it easy to gather basic information, conduct research, or craft a small article or message. You likely have your own method, but feel free to borrow this one.
The Method That Separates AI Beginners from Professionals: SMARTCHAIN
This is my approach to simple prompting.
It’s called SMARTCHAIN. (other methods exist!)
I know it sounds like something out of a corporate deck. It’s not. It’s a practical simple method that teaches you how to speak to AI clearly, so it gives you back what you actually need, not what it thinks you meant. And it’s an easy mnemonic so that I can work through it myself. Or you can save this mail or copy and paste it so it’s is readily available in a text expander or on your notepad or notion.
This isn’t about fancy hacks. It’s about thinking in layers. Structured thought, made visible in a prompt. That’s what SMARTCHAIN gives you.
Here’s how it works.
Situation
What’s going on? What are we solving? Most prompts are vague. They start in the middle of a thought. They assume the AI knows the context, your goal, and the stakes. It doesn’t.
Let’s say you want ChatGPT to write a job description. You type:
“Write a JD for a marketing director.”
That’s like walking into a meeting and saying, “Just talk.” No agenda. No tone. No goal.
A SMARTCHAIN prompt begins by anchoring the Situation:
“Our company is hiring a marketing director to lead global campaigns. We’re entering three new markets in the next 12 months and need someone who can own that expansion.”
Now the AI understands the world it’s operating in.
Mission
What do you want back—and in what form?
Be specific. Not “help me with this,” but:
“Write a 400-word JD in bullet points, ready for LinkedIn. Deadline: today.”
Clear task. Clear shape. Clear timeframe. You’re not delegating to a magic box. You’re giving an intelligent assistant a job.
Authority
Who’s the AI pretending to be?
You’ll get very different answers from a coach, an economist, or a sarcastic teenager. This layer makes the AI’s voice match your intent.
“You are a senior recruiter writing for a high-growth tech company.”
Or
“You are a witty ad copywriter who’s seen too many boring job posts.”
This one line shifts the tone, pace, and focus of the entire output. It’s not decoration. It’s direction.
Rules
Style and voice, made explicit.
Want British spelling? Short sentences? No passive voice? Say it.
“Use UK spelling. Make it punchy. Avoid buzzwords. No clichés.”
Rules aren’t restrictions. They’re shortcuts to taste.
When you’re training a junior team member, you give style guidance. Do the same with AI.
Tools
Add sources. Give it fuel.
“Use the 2024 LinkedIn Top Skills Report.”
“Refer to our company tone of voice doc.”
“Pull from this job ad: [paste link or text]”
Without tools, the AI invents. With tools, it builds.
This is especially useful for long tasks like reports, summaries, and competitive analysis. Feed the model your raw material. Then it doesn’t guess—it builds.
Checkpoints
How will you judge success?
This is where you get honest about expectations. Don’t assume perfection. Define what “good” looks like.
“JD must include salary band, benefits, three responsibilities, and a diversity statement.”
Make it measurable. Make it testable.
Even if you don’t tell the AI this part, you need it. Otherwise you’ll tweak forever, unsure what’s missing.
Handover
Hand it all over as one clean paragraph.
Now, combine everything above into a single, clean brief.
Example:
“We’re hiring a marketing director to lead expansion into three new markets. Write a 400-word JD in bullet points for LinkedIn. Deadline: today. You are a senior recruiter in a UK tech firm. Use British spelling, plain language, no buzzwords. Refer to the 2024 LinkedIn Top Skills Report. Must include salary, benefits, 3 key duties, and a diversity commitment.”
That’s the full SMARTCHAIN in one hit. Less guesswork. More signal.
Analyse
Slow down and think before prompting.
SMARTCHAIN isn’t about writing longer prompts. It’s about clearer thinking.
Ask yourself:
What’s the real task here?
Who will read this?
What would “bad” look like?
The act of structuring your prompt forces you to clarify your own goal. Half the time, you realise you weren’t clear, so how could the AI be?
Inspect
Don’t settle for “close enough.”
Check what the model gives you:
Did it follow the rules?
Did it match the tone?
Did it include what you asked?
If not, refine one layer and try again. Don’t rewrite the whole thing. Just tweak the right control.
Prompting is design. Inspect like a designer.
Next Steps
Ask what comes after.
If the job ad is done, what’s next?
“Now write an internal email to announce this role to staff.”
“Create 3 social media versions of the JD.”
“Draft 5 interview questions based on this brief.”
Layering leads to compounding. One good prompt becomes ten great outputs.
The Power Isn’t the Tool. It’s the Structure.
Anyone can type into ChatGPT. But not everyone knows how to structure thought into a prompt.
That’s what SMARTCHAIN does. It forces clarity. It makes the AI better by making you clearer.
It works for execs writing strategy. It works for marketers writing copy. It works for students writing essays. It works for you.
And once you try it, you won’t go back.
Good prompts don’t start with a question.
They start with structure.
Use SMARTCHAIN once and let me know if your output is improved.
Below is an example of one you could use for writing content on
CUT AND PASTE THIS INTO CHAT-GPT, GEMINI, CLAUDE, COPILOT AND REPLACE THE TEXT IN [SQUARE BRACKETS]
SYSTEM
You are a concise expert writer.
USER
We will use the ten-stage SMARTCHAIN method:
Situation • Mission • Authority • Rules • Tools • Checkpoints • Handover • Analyse • Inspect • Next steps.
Fill every CAPITAL placeholder before sending.
Leave no CAPITAL text in the final prompt.
Situation: [WHO OR WHAT IS THE PIECE ABOUT?]
Mission: [OUTPUT FORMAT, LENGTH, DEADLINE]
Authority: [EXPERT ROLE FOR THE MODEL]
Rules:
– British spelling
– Short direct sentences
– No clichés
– No passive voice
Tools:
– [LINKS, DATA, GLOSSARY - ONE PER LINE]
Checkpoints:
– [HOW YOU WILL TEST THE DRAFT]
Handover: Restate the brief in one paragraph, then wait for my “GO”.
Analyse: Think step-by-step. Outline three reasoning paths before drafting.
Inspect: Score against Rules and Checkpoints. Fix anything under 8/10.
Next steps: Suggest one follow-up prompt.
INSTRUCTION TO MODEL
Do not assume the brief.
If any CAPITALS remain, ask a single, specific question for each missing item and stop.
Once every CAPITAL is replaced, produce the ten SMARTCHAIN sections in order, exactly once.
END
Here is an example of me using this approach in researching AI Governance policy globally, so you can see the approach in practice.
As I said, this is simple and will help those at an early stage in prompting to use this. There are other methods, but this will improve your prompting significantly if you are just using AI with a single question.
An alternative approach for those using AI more regularly is LYRA.
LYRA
Lyra is a prompt optimiser that helps you decide on the best prompt for your goal.
Using AI every day? Try LYRA for prompt refinement. Copy the block below into ChatGPT and follow the instructions.
COPY THIS AND PASTE INTO CHATGPT
You are Lyra, a master-level AI prompt optimization specialist. Your mission: transform any user input into precision-crafted prompts that unlock AI's full potential across all platforms.
THE 4-D METHODOLOGY
#1. DECONSTRUCT
Extract core intent, key entities, and context
Identify output requirements and constraints
Map what's provided vs. what's missing
#2. DIAGNOSE
Audit for clarity gaps and ambiguity
Check specificity and completeness
Assess structure and complexity needs
#3. DEVELOP
Select optimal techniques based on request type:
Creative – Multi-perspective + tone emphasis
Technical – Constraint-based + precision focus
Instructional – Few-shot examples + clear structure
Complex – Chain-of-thought + systematic frameworks
Assign appropriate AI role/expertise
Enhance content and implement logical structure
#4. DELIVER
Construct optimized prompt
Format based on complexity
Provide implementation guidance
---
🔧 OPTIMIZATION TECHNIQUES
🏗Foundation: Role assignment, context layering, output specs, task decomposition
🧠Advanced: Chain-of-thought, few-shot learning, multi-perspective analysis, constraint optimization
Platform Notes:
ChatGPT/GPT-4: Structured sections, conversation starters
Claude: Longer context, reasoning frameworks
Gemini: Creative tasks, comparative analysis
Others: Apply universal best practices
---
🔄 OPERATING MODES
🔍DETAIL MODE:
Gather context with smart defaults
Ask 2–3 targeted clarifying questions
Provide comprehensive optimization
⚡BASIC MODE:
Quick fix primary issues
Apply core techniques only
Deliver ready-to-use prompt
---
🧾 RESPONSE FORMATS
➡️Simple Requests:
Your Optimized Prompt:
[Improved prompt]
What Changed: [Key improvements]
➡️Complex Requests:
Your Optimized Prompt:
[Improved prompt]
Key Improvements: [Prompt changes and benefits]
Techniques Applied: (Brief mention)
Pro Tip: [Usage guidance]
---
📝 WELCOME MESSAGE (REQUIRED)
When activated, display EXACTLY:
“Hello! I’m Lyra, your AI prompt optimizer. I transform vague requests into precise, effective prompts that deliver better results.
What I need to know:
Target AI: ChatGPT, Claude, Gemini, or Other
Prompt Style: DETAIL (I’ll ask clarifying questions first) or BASIC (quick optimization)
Examples:
“DETAIL using ChatGPT – write me a marketing email”
“BASIC using Claude – Help with my resume”
Just share your rough prompt and I’ll handle the optimization!”
---
🔁 PROCESSING FLOW
1. Auto-detect complexity:
Simple tasks → BASIC mode
Complex/professional → DETAIL mode
2. Inform user with override option
3. Execute chosen mode protocol
4. Deliver optimized prompt
Memory Note: Do not save any information from optimization sessions to memory.
END
You can see the original post here by Min Choi.
Stay Curious - and don’t forget to be amazing,
PS. Try SMARTCHAIN once this week and email me your best result.
Free your newsletters from the inbox: Meco is a distraction-free space for reading newsletters outside the inbox. The app has features designed to supercharge your learnings from your favourite writers. Become a more productive reader and cut out the noise with Meco - try the app today
If you enjoyed this edition of Box of Amazing, please forward this to a friend. If you quote me, or use anything from this newsletter, please link back to it.
If you are interested in advertising in this newsletter, please see this link