A real positioning diagnosis for a boutique agency feeling small next to enterprise competitors.
This is a trimmed sample. The full skill goes deeper — producing a Problem Brief, Rough Work artefact, and defended Final Deliverable with three approval gates.
Five moves. Three gates. No jumping to analysis.
The most common failure mode of brand work is solving the stated problem before diagnosing the real one. The skill is built to make that impossible. It runs a five-move loop and halts at three explicit approval gates so you stay in control of the work.
Find the real problem.
Captures the stated problem in your words. Forms a hypothesis. Routes the problem to the right area. Confirms the diagnosis back to you before going anywhere.
Pick the smallest toolkit.
Selects the right tactic, or the smallest combination, that produces clarity. Tells you the plan in plain language before running it.
Execute the tactics.
Pulls inputs from you in natural conversation. Runs the chosen tactics. Produces rough work with visual analysis. Halts so you can redirect.
Land on a recommendation.
Refines the rough work into a final recommendation. Leads with the call. Defends it with reasoning. Asks how you'll use the output.
Stress-test and ship.
Pressure-tests the recommendation. Confirms it holds under scrutiny. Ships the final deliverable in the format you'll actually use.
Gates 1, 2, and 3 require explicit approval from you before the next stage runs. "Continue," "go ahead," or "approved" qualifies. Silence does not.
A drawer for every brand problem.
Six diagnostic drawers. The strategist reaches into the one that matches the diagnosis. Each drawer holds a set of tactics. Running every drawer on one brand makes the brand weaker. The skill picks the right combination for your actual problem.
Competitor and market terrain.
Six tactics on finding alternatives, mapping where competitors are strongest, naming the human anxiety the category is built on, and breaking unspoken category assumptions.
When the brand does not know who it is really competing against.
Self-knowledge.
Six tactics on naming what only you do, surfacing weaknesses anonymously, learning what superfans love, and seeing where the market is actually pulling you.
When the brand does not know what it is actually working with.
Differentiation.
Ten tactics on compressing strategy into a three-part frame, naming what you do for the people who buy it, finding values that go against competitors, and narrowing the target.
When the brand cannot say what only it does or who it is really for.
Stress-test.
Twelve tactics on whether the strategy forces the business to change, whether it survives a stranger explaining it, whether subjective words can be stripped without losing meaning.
When the strategy looks good on paper but has not been pressure-tested.
Activation.
Ten tactics on assigning a strategy owner, turning strategy into consumer-facing message, mapping touchpoints, and making unbranded parts of the business expressive.
When the strategy exists but the business does not live it.
Decay and growth.
Six tactics on writing the document the brand defers to, measuring how outsiders see it, finding decisions competitors would never make, and taking the brand into culture.
When the brand is being copied, the market has shifted, or the next horizon needs claiming.
Three sequential outputs. Each one a gate.
You don't get a deliverable at the end. You get three deliverables along the way, each one earning its place by your approval. The strategist halts after each one.
The Problem Brief
After diagnosis. Before any tactic runs.
A concise statement of the real problem, the area it lives in, and the plan for solving it. Written in plain language. Halts the work until you confirm or correct.
The Rough Work
Visual analysis with the working direction.
An HTML artefact carrying the tactic outputs, at least one spatial artefact (position grid, competitor map, value chain), and a clearly labelled working direction. Halts until you confirm or redirect.
The Final Deliverable
The recommendation, defended.
The call leads. The defence follows. Format chosen by how you'll use it: docx to reference and share, slide deck to present, or live HTML to ship. Stress-tested before it leaves the room.
Inside Claude with the skill installed.
A short look at Gate 1 in action. The strategist refuses to start analysis until the real problem is named.
Built for the people doing the work, not watching it.
- Founders who want a strategist on call without retainer rates.
- Brand leads running internal strategy work and needing a senior second opinion.
- Freelancers and consultants who want a real methodology to apply with clients.
- Agency owners who want their own framework to run instead of borrowing a generic one.
- Anyone preparing to brief an agency, designer, or copywriter and refusing to do it badly.
- ×People who want a chatbot to write taglines and call it a brand.
- ×Anyone hoping Claude will replace doing the thinking themselves.
- ×Pre-product founders. You need something real to apply this to.
- ×People who want a logo. The skill outputs strategy, not visual design.
- ×Anyone uncomfortable being asked hard questions about their own work.