Offer Architect . Agency OS

What it does

Produces a complete offer architecture in ten parts: overview, avatar, dream outcome statement, core offer, value stack with assigned values, risk reversal, decision accelerators, the offer name, a priced recommendation with rationale, and a value equation scorecard rating each variable out of ten. The strategic engine is Alex Hormozi's value equation, but the vocabulary adapts to your positioning tier, so a luxury business gets exclusive inclusions where an info product gets bonuses. It reads your ICA first so you never re-explain your customer.

Say this to start

This skill has no button. You start it by saying what you want. Any of these will do it:

/build-offer

> offer architect

> build my offer

> value stack

> irresistible offer

> pricing and guarantees

When to reach for it

Build 3, straight after the ICA is done
Your current offer is not converting and you do not know which part is broken
You suspect you are undercharging
Packaging agency services into retainer, project or hybrid pricing
Building a client's offer inside their workspace
Your guarantee is really just a refund policy

When NOT to use it

If you actually want	Use this instead
Defining who the offer is aimed at	`ideal-customer-avatar (run it first)`
The sales page, funnel or email sequence that sells the offer	`funnel-architect, with legendary-copywriters for the copy`
Converting inbound leads on sales calls	`sales-conversion-engine`
Your agency's own outbound client acquisition pipeline	`agency-sales-system`

Before you start

What you need	Why
`ica-output.md` for this product	the skill checks for it before Phase 1 and warns that an offer without a specific avatar will be generic. It offers to run the ICA first	Required
`verified-claims.md` with real, consented proof	no registry entry means every number, result, testimonial and credential is stripped out and replaced with a proof-gap marker	Required
Your current offer and price point, if you have one	Phase 1 asks for it directly as the anchor for the rework	Optional
Your delivery model: done-for-you, coaching, course, SaaS or hybrid	it determines the vocabulary, the container word and the pricing structure	Required
The single most common reason prospects say no	the problem mapping stage is built around dismantling it	Required
Your team structure and delivery capacity	a solo operator gets one well-anchored offer instead of tiers, since a premium tier that sells more of your own time does not scale	Optional

How it runs

Prerequisite and context loadChecks that this product has an ICA and offers to run it if not. Reads your owner profile, voice profile, avatar and any prior offer work, then applies past feedback from learnings.md.
Market research passResearches competitor pricing tables, value stacks, guarantees and bonus structures, what offer shapes are currently converting in your category, and how buyers there evaluate offers. Asks whether you want to name your top three competitors.
Phase 1: discoveryFour questions asked conversationally: what you do and who for, what you sell now and at what price, how you deliver, and the number one reason prospects do not buy. Then it names your positioning tier (luxury, premium, professional or accessible) and asks you to confirm it.
Components 1 to 3Checks the market has pain, purchasing power, targetability and growth. Applies the value equation: dream outcome times perceived likelihood, divided by time delay times effort. Then lists every problem between the customer and their outcome and designs a solution for each.
Components 4 to 6Builds the value stack so the core offer feels like a fraction of what is included, designs the risk reversal (unconditional, conditional, or both stacked), and sets authentic scarcity or urgency. Manufactured scarcity is rejected outright, especially at premium tiers.
Components 7 and 8Names the offer using the MAGIC formula (magnetic reason, avatar, goal, interval, container word) with a container word that fits your tier, then sets the price with a rationale tied back to the value equation.
Grounded claims checkRuns two passes before anything is presented. A linter catches numbers and anecdote-smell, then a grader reads the whole draft for any specific a reader could ask says who about. Nothing is shown until both come back clean.
Phase 3: output and gateCompiles the ten-part offer, runs a copy grader over the customer-facing sections against your ICA, then executes the six-step closeout: auto-score, .docx, Google Drive if available, dashboard, slide opt-in, build overview.

What you get

[target]/offer-output.md, the full ten-part offer architecture in schema form
[target]/positioning.md, your market positioning, differentiation, pricing tier and competitive angles
A value equation scorecard rating each variable one to ten with improvement notes
A .docx version of the offer document
A dashboard section and an optional slide deck
A market intelligence section appended to market-intel.md

Honest limits

Read this before you rely on it

It will not let you use a number, result, testimonial or credential that is not in verified-claims.md. With an empty registry the offer comes out generic with proof-gap markers, and that is deliberate.
The claim check is two-tier and blocking. If the grader finds a fabricated specific the draft is not shown to you until it is grounded or genericized.
It gives you a recommended price, not a validated one. The market decides whether the number holds.
It is deliberately blunt. When your offer is weak, your guarantee is really a refund policy, or you are undercharging, it says so.
Output quality tracks input quality. A thin avatar produces a thin offer, and the market research pass writes an error stub and continues if its research tools time out.

Where people go wrong

The mistake	Do this instead
Padding the value stack with irrelevant bonuses	Every element must remove a specific obstacle on the path to the outcome. If it does not accelerate the result, cut it. Padding makes prospects question your judgement.
Offering satisfaction guaranteed	That is a refund policy in costume. Make it conditional and specific: if you do X and do not see Y by this date, we do Z.
Inventing spots left to create urgency	Use a real constraint such as your calendar, onboarding capacity or cohort size. If there is no natural limit, use a deadline instead of fake scarcity.
Stating a price with nothing to compare it to	Anchor first: the cost of the problem, the value of the outcome, or a higher-priced alternative. Then present the price as a fraction of that.
Building three tiers as a solo operator	When your own delivery capacity is the binding constraint, one well-anchored offer plus at most one down-sell beats tiers. A premium tier that sells more of your time shackles you.
Reaching for the most impressive number you can think of	Use the smaller result you can document over the bigger one you cannot. Anything unsourced gets flagged and stays out of client-facing copy until it is verified.

Worth knowing

The value equation has four levers and most people only pull one. Making the dream outcome bigger is the crowded move. Driving the bottom half toward zero, less waiting and less effort required from the buyer, is what the strongest companies compete on.