Using the frameworks
Five steps, and the first output in about half an hour. You never edit a framework or fill anything in first — the interview does that for you.
Find a framework
Search or browse the library for the challenge in front of you. Each framework is built for a specific situation rather than a general topic.
Copy it
One click copies the whole framework. You never edit it or fill anything in first.
Paste it into your AI
ChatGPT, Claude, Gemini, or whatever you already use. The framework carries its own instructions.
Answer the interview
The AI asks one question at a time to draw out your context. This is where the quality comes from.
Refine the output
Read it critically and push back through follow-up conversation. Treat the first output as a starting point.
Check your privacy settings first
If you are sharing sensitive business information, confidential data, or anything that should not be in the public domain, check your AI tool's privacy settings before you start. Most providers offer options to prevent your conversations being used for model training. In ChatGPT, disable "Improve the model for everyone" in Settings → Data Controls. In Claude, check your workspace settings.
Why structure beats one-liners
A simple request like “Write me a business case” will give you something generic. The AI has no context about your situation, your constraints, or what good looks like in your organisation. It fills the gaps with assumptions, and those assumptions are usually wrong.
The frameworks work differently. They establish credibility through a practitioner role, define the specific task, request the context that actually matters, and specify a structured output that covers what experienced operators know to include.
What's inside a framework
Practitioner role with credibility
Opens with a senior role and specific experience. This shapes how the AI approaches the problem, drawing on patterns from similar situations rather than generic advice.
Context inputs that matter
Specifies what information the AI needs to give you a tailored answer. These are the inputs an experienced consultant would ask for before giving advice.
Structured output specification
Defines what the deliverable should cover. The sections reflect what practitioners know to include, the things that get missed when you are rushing or new to a domain.
Interview instruction
Tells the AI to ask questions before producing output. This ensures the response is based on your actual situation, not assumptions.
About the interview
Every framework ends with an instruction for the AI to interview you before producing output. This is where the real value lies. The questions force you to think through your situation properly, and they give the AI the context it needs to be genuinely useful.
Think of it as a conversation with a knowledgeable colleague who asks the right questions before jumping to conclusions. The process of answering often clarifies your own thinking.
If you are time-pressed and already know exactly what you need, delete everything after the horizontal line (---) and provide your context in the same message. You will get a faster response, though it will be less tailored.
Speak your answers
The interview draws substantial context out of you, and typing detailed responses slows you down. Voice-to-text lets you answer naturally, which is faster and usually produces richer context.
We recommend Wisprflow which works across any application on your computer. Use this link for a free month to try it out.
What one looks like
You are a Programme Director who has kicked off dozens of complex initiatives across retail, financial services, and private equity portfolio companies. You understand that the first 30 days of a programme set the trajectory for everything that follows.
You are helping structure the discovery phase of a new programme.
I will share:
- Programme objectives and success criteria
- Key stakeholders and their interests
- Known constraints and dependencies
- Timeline expectations
- Available resources and budget parameters
Please provide a discovery phase plan covering:
1. **Stakeholder Landscape**
- Key decision-makers and their priorities
- Influencers and potential blockers
- Communication preferences and cadence
2. **Current State Assessment**
- Information sources to review
- Interviews to conduct
- Data to gather
[... continues with structured sections ...]
The quality of your discovery phase determines whether you spend the programme fighting fires or building momentum.
---
Before producing any output, interview me one question at a time to gather the context you need. Ask only one question, wait for my response, then ask the next. Continue until you have enough information to produce a high-quality, tailored output.A thought partner, not a replacement
The goal is not to have AI do your job. It is to give you a thinking partner who can help you structure problems, challenge your assumptions, and produce first drafts you can build on.
AI is good at structure, synthesis, and generating options. You bring judgment, context, and accountability. The combination is more powerful than either alone.
Ready to try one?
Browse the library to find a framework for the challenge in front of you.
Looking for the Boardroom?
Frameworks are for one problem at a time. The Boardroom installs a standing exec team inside your AI assistant — different setup, same five minutes.
Setting up your Boardroom