Sign inGet started
Skills / Product-management

Shape Up — Collaborative Shaping

Appetite is different from an estimate. Estimates start with a design and ask how long. Appetites start with a time and ask what fits.

Shape Up Collaborative Shaping is the upstream phase of Ryan Singer's Shape Up methodology, where raw product ideas are transformed into well-defined Pitches before any team commits to building. Shapers fix a time budget (the 'appetite'), sketch solutions at low fidelity using breadboards and fat marker sketches, and explicitly call out rabbit holes — all before the work reaches the betting table. The result is shaped work with the right scope, the right risks named, and enough structure for teams to evaluate and build without mid-cycle surprises.

By Ryan Singer · Free
Specimen 01 · Live diagnosisShape Up — Collaborative Shaping
Input

“We've been thinking about a project status dashboard. Managers keep asking project leads for updates and it's interrupting everyone.”

Diagnosis
Let's shape this before it becomes a spec.
Full transcript ↓
Calibrated referenceagent-skills.ai
The gap

Set the appetite, sketch the breadboard, name the rabbit holes — then bet.

Shape Up treats shaping as a separate, disciplined phase from building — work is defined upstream before any team commitment is made. The process follows a fixed sequence: set the Appetite (a 2-week small batch or 6-week big batch time budget) before touching the solution; then define the problem in situation-specific terms rather than abstract requirements; then sketch solutions at low fidelity using Fat Marker Sketches and Breadboards — text-based flow diagrams showing Places (screens/menus), Affordances (buttons/fields), and Connection arrows, without any visual design. Shapers then name the Rabbit Holes (risky technical unknowns that could derail the cycle) and compile everything into a structured Pitch document comprising Problem, Appetite, Solution, Rabbit Holes, and No-gos. The Pitch goes to the Betting Table, where shaped work competes for a cycle slot on its merits.

The problem

Product teams waste cycles on work that was never fully shaped — handed off as vague tickets or over-specified mockups, both of which force builders to make product decisions mid-cycle when it's most expensive. The default process (estimate first, spec second, build third) bakes in scope ambiguity and technical unknowns that surface only after the team is already committed. Shape Up's shaping phase solves this by establishing a dedicated upstream phase where solutions are defined at the right fidelity: rough enough to keep implementation options open, specific enough to prevent rabbit holes from swallowing the cycle.

The solution

Stop handing your team ambiguous problems or over-specified mockups. Walk away with a Pitch that names the appetite, shows the breadboard, calls out the rabbit holes, and gives the betting table everything it needs to make a confident cycle decision.

You bring
  • A raw product idea or customer pain you're considering addressing
  • A rough sense of how important the problem is relative to your team's time
  • Any known constraints, prior solutions, or technical concerns
  • The specific user type or persona experiencing the problem
You get
  • A complete five-section Shape Up Pitch (Problem, Appetite, Solution, Rabbit Holes, No-gos)
  • At least one Breadboard diagram showing Places, Affordances, and Connection flows
  • An explicit Rabbit Holes list with recommended resolutions or No-go calls
  • A narrowed Problem statement grounded in a specific situation, not abstract requirements
In action

Watch the methodology work.

Three specimens from a single real session: the same situation, unaided and calibrated, the full transcript, and the skill answering live in the channel where the work happens.

Fig.01The same situation, two ways
Specimens A · B
Specimen AVague
Unaided

“A PM drops a Jira ticket: 'Add project status dashboard — managers want visibility.' No problem definition, no appetite. A week into the six-week cycle, the team is debating whether to build a custom reporting engine or integrate a third-party tool, the backend engineer discovers the data model doesn't track project health in any structured way, and what started as 'a dashboard' has tripled in scope. The cycle ends with a half-built feature, a frustrated team, and a PM wondering why it's always like this.”

×Vague — nothing concrete to act on.
Specimen BCalibrated
With Shape Up — Collaborative Shaping

“A 45-minute shaping session produces a Pitch with a clear 2-week appetite, a breadboard showing a Digest Email flow with Project Settings toggle, an explicit rabbit hole (data model gap — spike before betting), and a No-go fallback (manual digest if auto-generation isn't feasible). The betting table reviews it in ten minutes, assigns a backend engineer to spike the data question that week, and schedules the feature for the following cycle with full confidence in the scope.”

Specific, grounded, and ends with one clear next step.
Fig.02A real session, unedited
In-app · assistant
Shape Up — Collaborative ShapingClaude
Shape Up — Collaborative Shaping
Let's shape this before it becomes a spec. Start with the problem, not the solution — describe a specific moment where someone runs into friction. Not 'we need X feature,' but the actual situation: who's there, what are they trying to do, and what makes it painful or slow right now?
We've been thinking about a project status dashboard. Managers keep asking project leads for updates and it's interrupting everyone.
Let's find the real problem — 'dashboard' is already a solution. When a manager asks for status, what actually happens next? Walk me through the moment: the manager asks, and then what does the project lead have to do?
And before we sketch anything: how much is solving this worth to you? Is this a 6-week big batch — a real reporting system — or a 2-week small batch that just reduces the interruptions?
The lead has to stop working, pull up three different tools, piece together a picture, and write a Slack summary. It happens a few times a week. And yeah — two weeks feels right. We don't need a perfect system, just fewer interruptions.
Good. Problem: project leads lose focused work time assembling status summaries that could be automated. Appetite: 2 weeks.
Let me breadboard a lightweight solution without designing it yet:
Project Settings → [Digest toggle: on/off] → [Frequency: daily / weekly] Digest email → Project health snapshot → [Open project link] Digest email → [Send now] → Delivery confirmation
One question before we go further: does the digest pull data automatically, or does the project lead write anything into it? That answer changes the scope entirely.
It should be fully automatic — if they have to write anything, nobody will actually use it. But now I'm wondering... we don't really have a structured 'project health' status anywhere in our data model.
That's a rabbit hole — and you just surfaced it before it could wreck the cycle. Let's name it explicitly: Can the current data model generate a meaningful health snapshot without manual input from the lead? This needs a technical spike before the pitch goes to the betting table, otherwise you're betting on an assumption.
Two paths: (1) Mark it as a rabbit hole and have a backend engineer investigate before the bet. (2) Make auto-generation a No-go for this batch and scope the digest to manual one-click summaries — the lead fills in two fields and hits Send. Which do you want to take to the betting table?
Let's mark it as a rabbit hole and have the backend lead check. If the data isn't there, we fall back to the manual version. I can actually see the full pitch taking shape now — this is way more structured than how we usually approach new features.
Reply to Shape Up — Collaborative Shaping
What the skill is doing
01Appetite Setting
02Problem Definition Sharpening
03Breadboard Construction
Fig.03The same skill, in the channel you already use
WhatsApp · live
9:41
SUShape Up — Collaborative Shapingtyping…
Messages are end-to-end encrypted. No one outside of this chat can read them.
Today
We've been thinking about a project status dashboard. Managers keep asking project leads for updates and it's interrupting everyone.9:14✓✓
Let's shape this before it becomes a spec.9:14
Start with the problem, not the solution — describe a specific moment where someone runs into friction.9:15
Not 'we need X feature,' but the actual situation: who's there, what are they trying to do, and what makes it painful or slow righ…9:15
Message

The same skill, where the work happens.

No new app to learn. The methodology runs over the WhatsApp Business API, so the answer lands as a reply in the thread you’re already in — same rigour, zero context-switch.

Reads the situation, names the pattern, returns one concrete next move.
Delivered in seconds, inside a conversation that already exists.
Specimen · WhatsApp Business API · live
Capabilities

What it does, specifically.

Each capability is a distinct move drawn straight from the source methodology — not a generic assistant guessing.

CapabilityC-01

Appetite Setting

Reframes your scope conversation around time as a strategic constraint rather than a byproduct of estimation. Instead of asking how long a design will take, you first decide how much the problem is worth to your business — then let that appetite define what the solution can be. This prevents runaway scope and forces honest prioritization before any design work begins.

Based on Singer's core inversion in Shape Up: appetite is not an estimate but a deliberate business decision about value — 'how much is this problem worth to us?' — which constrains the solution rather than the solution constraining the timeline.
CapabilityC-02

Problem Definition Sharpening

Transforms vague feature requests into situation-specific problem statements by forcing a real-world framing: who is in the moment, what are they trying to do, and what makes the current approach painful or broken. Specific problems produce bounded solutions; abstract requirements produce endless scope debates.

Based on Singer's emphasis in Shape Up on defining problems in terms of specific situations rather than generalized needs — the narrower the problem framing, the more clearly the right solution presents itself.
CapabilityC-03

Breadboard Construction

Produces a text-based flow diagram — in the style of an electrical breadboard schematic — capturing Places (screens, menus, modals), Affordances (buttons, fields, links), and Connection arrows showing navigation. The deliberately low-fidelity format avoids premature visual design decisions while still communicating the interaction structure clearly enough for a betting table to evaluate.

Based on Singer's breadboarding technique from Shape Up: using the electrical breadboard metaphor to sketch user flows at the level of places and affordances without committing to visual design, preserving room for builder judgment during implementation.
CapabilityC-04

Rabbit Hole Excavation

Systematically interrogates your proposed solution for technical unknowns, edge cases, and scope traps that could derail a cycle mid-stream. Each rabbit hole is named explicitly and assigned a resolution path: spike it before betting, make it an explicit No-go, or constrain the solution to avoid it. Surfacing these before the bet is far cheaper than discovering them during building.

Based on Shape Up's dedicated rabbit holes step — Singer treats naming rabbit holes as a required pre-betting activity, arguing that the shaper's job is to resolve risks while the solution is still clay, not mid-cycle when the team is already committed.
CapabilityC-05

Pitch Document Assembly

Guides you through the complete five-section Pitch structure — Problem, Appetite, Solution (with breadboard or fat marker sketch), Rabbit Holes, and No-gos — producing a formatted document ready for betting table review. Each section is grounded in the shaping work done upstream, so the pitch is compiled from decisions already made rather than written cold.

Based on the canonical Pitch format from the Shape Up book, which Singer defines as the artifact that carries shaped work to the betting table — complete enough to evaluate, rough enough to preserve builder autonomy.
Tested

Graded before it shipped.

Every skill is scored against independent scenarios for methodology fidelity before it goes live — not vibes, a rubric.

What it produces
OutputD-01

Shape Up Pitch Document

A structured five-section document (Problem, Appetite, Solution, Rabbit Holes, No-gos) formatted for a betting table decision. Contains the breadboard diagram, the scoped appetite, and all known risks explicitly called out — everything a team needs to decide whether this work deserves a cycle slot.

OutputD-02

Breadboard Diagram

A text-based flow diagram showing Places (screens, menus), Affordances (buttons, fields, links), and Connection arrows — the Shape Up way to communicate solution structure without committing to visual design. Gives builders enough structure to start without over-specifying how it should look.

OutputD-03

Rabbit Holes Register

An explicit list of technical unknowns, edge cases, and scope traps identified during shaping, each with a recommended resolution: spike before betting, mark as No-go, or constrain the solution to avoid it. Prevents mid-cycle surprises by surfacing risks while the work is still shapeable.

The source

Grounded in the original work.

Every answer traces back to a real source and the practitioner who wrote it — not a secondhand summary. Here is the source of record.

Source authorA-01

Ryan Singer

Ryan Singer is the creator of the Shape Up methodology and a longtime product strategist at Basecamp (37signals), the company behind the Basecamp project management tool and Hey email service. He developed Shape Up through years of direct product development work at 37signals, documenting the full methodology in the freely available book Shape Up (basecamp.com/shapeup). Singer has shared the approach through the Shape Up podcast and multiple appearances on the Rework podcast.

Status · Inspired by Ryan Singer’s work — not yet claimed. Are you Ryan Singer?
Primary sourceS-01

Shape Up: Stop Running in Circles and Ship Work that Matters

by Ryan Singer

Product strategist at Basecamp (37signals); creator of the Shape Up methodology; author of Shape Up (basecamp.com/shapeup)

Read the original ↗
Citationbasecamp.com
Get started

Put it to work.

agent-skills runs as MCP — connect it once and Shape Up — Collaborative Shaping works in Claude, ChatGPT, Cursor, or any MCP-compatible client. Sign in to see your config and accept the terms.