The Clear Edge AI Agent Setup Guide: Install, Run, and Diagnose Your Business Constraint

What Is the Clear Edge Diagnostic Stack?

Every Clear Edge subscription gives you three tools that work together as one system.

The AI Diagnostic Session is the execution layer — a structured diagnostic file you load into any AI you already use. It reads your situation, identifies your real constraint, and walks you through your next move step by step. It calibrates to your income band, business stage, and context. No complicated setup — a single upload and it’s ready.

The PDF is the depth layer — detailed notes, examples, case studies, and the full operational structure behind the framework. Built so AI can read the entire system and work from it, and so agencies and consultants have a complete reference when applying it to clients. You don’t need to read it cover to cover — use it as a reference, go deep when you have the time, or let AI pull exactly what you need in minutes.

The Audio Keynotes are the listening layer — key insights from the article and framework distilled into focused audio. Built for commutes, review sessions, or anyone who absorbs better by listening. Not a recording of the article — a curated extraction of what matters most.

Together, they form one system — the AI session for fast execution, the PDF for depth, and the audio for on-the-go reinforcement. All three work on your own business or with clients.

Who This Stack Is Built For

Solo consultant who needs a sharper, constraint-first advisor on call
Agency owner who wants a repeatable diagnostic they can run across clients
Fractal CEO or operator running multiple bets and needing a single view of constraints
Digital creator turning attention and audience into a more deliberate business system

How to Install an AI Agent on Claude, ChatGPT, and Gemini

Unzip Your Agent File

Your download is a .zip file. Unzip it first — inside you’ll find two files:

The agent file (.md) — this is what you upload to your AI
The README file (.md) — this is your setup and reference guide

Keep both. You’ll need them.

Claude (Anthropic) — Recommended

Most precise for business diagnostics. Easier to manage files, track progress, and execute across sessions.

Skills: Claude.ai → Customize → Skills → Add new skill → upload the agent.md file→ name it → save. Available across all your Claude conversations.

ChatGPT (OpenAI)

Custom GPT: chatgpt.com/gpts → Create → Configure → upload the agent.md file in the Knowledge section → name it → Save

Gemini (Google)

Gem: gemini.google.com → Gems → New Gem → Add files for your Gem to reference → upload the agent.md file → name it → save.

Once the agent is loaded, you can see everything built into it — the behavioral framework, diagnostic tools, fast-track protocols, and tracking structures, all in one place. That’s the entire install. Your agent is now ready to use.

The Clear Edge doesn’t collect any of your personal data. Everything runs inside your own model, on your own account.

Note: Platform interfaces change frequently. The steps below reflect current setups but may look different by the time you read this. The goal is always the same: find where your AI platform accepts a custom agent, skill, or GPT — then upload the agent file there.

What to Do Before Starting Your First Session

Every agent comes with a README file. Open it before anything else. It contains install instructions for your platform, the trigger phrase to start your session, a summary of what the agent covers, and a checklist of what to have ready.

Common triggers across all Clear Edge agents:

Start a session: “Let’s run the framework.”
Resume: “Resume” — picks up where the last session ended
Show templates: “Show templates” — displays all available output templates
Build tracking files: “Build my tracking files and documents”
Show dashboard: “Show my dashboard” — pulls current metrics and session history

Your agent’s exact trigger phrase is in its README. The patterns above work across all agents.

Model Selection

Model selection matters. Different models — and different versions of the same model — respond differently. Lower-capability models may skip steps or miss nuance. For business diagnostics, use the latest available Opus model for the sharpest results.

Onboarding

On Claude, open a new conversation, press the / button, select the agent from the menu, then say, “Let’s run the framework.”

First-time setup takes under 30 seconds and only happens once. The agent will ask:

Your name
Business type — Agency, Service Business, Solo Consultant, Fractal CEO, or Digital Creator
Monthly revenue band — to calibrate response depth and protocol precision
Currency and preferred language

The session is flexible — you can ask the model to respond in any language, including Arabic, Chinese, Bahasa, or any other, and it will adapt. Income bands and business model thresholds are benchmarked in USD as the standard reference. If you work in a different currency, the model will convert and calibrate accordingly — just let it know.

If you’ve used this or any other Clear Edge agent before, the model may pull from prior context and continue where you left off. This is especially useful when working through layered constraints or ongoing problems.

How the AI Identifies Your Business Constraint

The agent runs a structured diagnostic sequence. The goal is to identify your real constraint — not the symptom, but the root cause — and build a specific, actionable plan around it.

It will ask questions like:

What’s your current monthly revenue and how long have you been at this level?
Where is the bottleneck — pricing, capacity, lead volume, conversion, or delivery?
What have you already tried, and what happened?
What does your current offer or service look like, and who is it for?
Score your activities’ revenue output from 0 to 10.

Answer precisely — real data beats estimates, but a rough number is fine. The model isn’t looking for perfection; it’s looking for patterns. Talk to it like a sharp advisor who already knows the framework — no jargon, no formality. If something doesn’t land right, just reset the conversation and start again.

Stay on topic. Going off-topic causes the model to drift.

Once the diagnosis is locked in, you and the agent build your next move together — specific to your constraints, with examples and step-by-step guidance. If you have multiple blockers, add them as notes. If you need more clarity, just ask — the agent can break anything down further or walk you through it step by step.

The work doesn’t stop at the diagnosis. The agent helps you track progress toward closing the constraint — daily or weekly — measuring whether the steps are being applied and whether they’re moving the numbers. Identifying the constraint is the start. Applying the fix, tracking the outcome, and iterating until the ceiling breaks is the goal.

Tracking Files

At the end of every session, ask the agent to build your tracking files and documents. Some models do this automatically when execution mode is selected during onboarding — others skip it. Don’t wait to find out. Just ask.

As your inputs grow, the agent updates those files and builds a running record of your constraints, decisions, and progress. If you have external tools connected — Google Sheets, Notion, email, or others — the agent can write directly to them. Just ask: “Connect to [tool name]” and it will guide you through the setup. If no tools are connected, the agent creates the files locally so you always have your data.

A note on AI: The Clear Edge materials are designed to sharpen your thinking and judgment, not replace them. The output is only as good as what you put in — be precise, be specific. Always cross-check the diagnosis against your actual situation before acting on it. AI can make mistakes. You’re the operator. The final call is always yours.

How to Use a PDF With AI for Business Diagnosis

The Clear Edge PDFs are structured to work dynamically with AI — the content responds to your specific situation, not a generic one-size-fits-all answer. Load the PDF into your AI and have a conversation with it. Newer PDFs include built-in prompt instructions. Older ones may not, but all work with the same four modes.

The Four Ways to Use Any Clear Edge PDF With AI

Extract — Understand the core framework fast. Key signals, one example per section, no filler. Use this first.

Apply — Bring your real situation. The model diagnoses first, then gives you the right protocol for your case — not a template answer.

Elicit — The PDF interviews you before it diagnoses. Useful when you’re not sure how to describe your situation or where to start.

Stress-test — Challenge the framework against your specific constraints before you commit. Find where it breaks, and what to adjust.

Copy-Paste AI Prompts for Business Diagnosis (With Examples)

Copy and paste directly into your AI with the PDF loaded.

Extract — Understand It Fast

No situation needed. Use this to compress the full framework before doing anything else.

Summarize this PDF's core framework, the key signals/rules, and one example per section. Compress to what I'd need to understand it in 5 minutes. Skip filler.

Apply — Run It on My Case

Bring your real situation. The model diagnoses first, then prescribes.

Diagnose my situation with this PDF, then give me the right detailed protocol for it. Use only what I provide — no invented figures. Diagnose before prescribing; flag gaps, mark assumptions.

My situation: Freelance web developer, mostly small business sites. I charge $100/hour. Billing maybe 75 hours/month but spending another 30+ on sales, admin, and revisions I don't bill. Around $32K/mo revenue across a rotating set of projects. No idea what my real margin is. I close almost everything I quote. I've used hourly for 3 years and never moved off it. Competitors seem to do flat project pricing but I don't know their numbers. Burning out doing unpaid admin and want to stop trading hours for money.

Elicit — Let It Interview You

Start vague. The PDF will ask the questions it needs before diagnosing.

Diagnose my situation with this PDF, but interview me first — ask the questions you need to fill the framework, then diagnose. Use only my answers — no invented figures, and flag anything I can't answer.

Honestly not sure how to explain it. I do social media management for a handful of local businesses, charge somewhere around $1,200/month each, and it feels too low but I can't tell. Not great with the numbers side.

Stress-test — Find Where It Breaks

Use this before committing to any protocol.

I've understood this PDF's framework and how it applies to me. Now challenge it: where does it break down for my situation, what is it missing or oversimplifying, and what would a smart skeptic say before I commit? Then tell me what to adjust to make it actually fit. Use only what I've provided — no invented figures.

My situation: Business coach, $1,200/mo, 15 clients, all from referrals who hired me because I'm affordable. The PDF says raise 300% and accept churn. But my entire pipeline IS those referrals — if I reprice and they leave, my lead source leaves with them. I have no other acquisition channel and 2 months of runway. I can't just "accept the churn." Where does this framework break for me?

For tracking and measurement, ask:

Build my tracking files and spreadsheets so I can track this constraint daily/weekly/monthly.

You’re Ready

You now have everything you need. The agent identifies the constraint, the PDF gives you the depth to understand it, and the audio keeps it sharp between sessions. The prompts above turn all the data into action.

This gives you an entire constraint breaking cycle:

Diagnose → Plan → Stress-test → Execute → Measure → Improve

Load the agent, run the session, apply the PDF, reinforce with the audio. The constraint is clear — now it’s on you to close the gap and move to the next ceiling.

Report a Problem or Leave Feedback

Report a Problem: clrdg.link/report
AI Session Feedback: clrdg.link/copilot-feedback

The Clear Edge OS — Built for six-figure operators who execute.