---
name: scope-and-brief
description: Turn a big, vague idea into a one-page brief you can decide on and hand off. Use when someone wants to automate, build out, restructure, launch or "set up" something that is bigger than one task, or asks to scope an idea before spending time or money on it.
---

# Scope and Brief

Most "can you do this for me" asks are not one thing. "Automate my marketing" is a multi-step build wearing the costume of a single task. Nobody can make a real decision about something they can't see the shape of.

This skill makes the shape visible: the real steps, the real inputs, the real cost. Then it hands the decision back to the user.

It does not do the work. It makes the work decidable and plannable.

## Posture

- Help the user see clearly. Never reject the ask.
- Ask one question at a time. Short sentences, plain words.
- Push back only when an answer is too vague to plan from.

## Steps

### 1. Name the real shape

Before anything else, say what the thing actually is. If it is more than one task, say so plainly:

> "That's not one task. It's a multi-step build with several pieces that depend on each other. Before you decide whether to take it on, let's map what it involves."

### 2. Read capability

The same idea needs a different plan for a beginner than for someone who has done it before. Ask, lightly:

> "Have you or your team done this kind of work before, or is it new ground?"

### 3. Find the constraints

Ask what is actually scarce right now: time, cash, a key person, attention. Ask what can't change: budget caps, compliance rules, deadlines, systems they're stuck with. Price everything that follows in these constraints, not in the abstract.

### 4. Map the progression

Lay out the steps in the order they have to happen. Name what depends on what, what inputs or connections each step needs, and where the unknowns are. Be concrete: "Step one needs the customer list exported from the CRM" beats "first, set up the foundation."

### 5. Write the brief: two layers, one document

**Decision layer (for the user, deciding now):**
- **What was asked:** the idea, stated plainly
- **What it actually is:** the real shape, the multi-step reality
- **Constraints in play:** what's binding
- **The investment:** time and money, honestly estimated. Ranges are fine. False precision is not.
- **What it costs to do it now:** what gets paused or starved to make room
- **Capability read:** can they execute this as-is, or is there a learning step first?

**Handoff layer (for whoever does the work next):**
- **The steps, in order,** with dependencies named
- **What's needed at each step:** inputs, access, people
- **The next concrete action,** small enough that someone can start without a meeting, with a suggested owner
- **Enough context** that the person doing the work isn't blocked asking why, what or from whom

Always write both layers, even if the user is leaning toward "not now." They can't decide honestly without seeing the cost.

### 6. Route the decision

Help the user pick one path:
- **Pursue now:** the brief becomes the founding document for the work.
- **Defer:** it's real, but not now. Name what would trigger a revisit.
- **Route:** it belongs somewhere else (a different person, project or tool). Say where, and what context that place needs.

End by writing the decision down in one line: what they chose, and why.

## Did the brief work?

Both answers must be yes:
1. **Can the user make a go/no-go call from it?** If not, the decision layer is too thin. Usually the cost estimate or the capability read.
2. **Can they hand it to someone who can move without a meeting?** If not, the handoff layer is too thin. Usually the inputs per step or missing context.

## Failure modes

- **Gatekeeping.** "Your ask isn't specific enough" means the skill failed. Show the shape. Hand back the decision.
- **Skipping the capability read.** A plan that ignores who's doing the work is a generic plan.
- **One layer instead of two.** Decision framing with no handoff is half a brief. Steps with no decision framing is the other half.
- **False precision.** "$4,200 and 31 hours" is a lie dressed as rigor. Use honest ranges and name the unknowns.
- **Doing instead of scoping.** If it starts building the thing, it has left its lane.
- **No decision at the end.** A brief with no path chosen is just more unresolved context.

---
Built by Michael Gardon · Chief AI Accelerator · accelerator.rejoincoaching.com
