scope-cutter - SKILL.md Agent Skill

name: scope-cutter description: Use when the user is deciding what to cut from an onboarding flow, MVP scope, or feature list. Triggers on phrases like "cut scope", "what should we cut", "required vs optional", "progressive disclosure", "trim onboarding", "what's blocking activation", "the onboarding is too long". Implements the Required-vs-Helpful scope-cutting heuristics from Ch 3 and Ch 4 of Building Rocketships.

Scope Cutter

What this does

Runs Ch 3/4's Required-vs-Helpful scope cutting on an onboarding flow or feature list. Returns a two-column allocation (Required vs Helpful) with rationale per item, plus the progressive-disclosure nudge sequence for everything that gets moved to Helpful. Targets the book's rule: cut the required-onboarding list in half.

When to use

Trigger when the user:

Has an onboarding flow with more than 4–5 required steps
Is shipping an SLC and needs to narrow scope (pairs with slc-or-mvp)
Has a "required fields" list on a form longer than feels reasonable
Is debating whether a feature belongs in v1 or v2
Has measured activation drop-off concentrated in onboarding

When NOT to use

The user is asking about the aha moment itself — that's aha-mapper, which feeds this skill its definition of "required"
The user is asking about tier feature segmentation (free vs paid) — that's reverse-freemium-design
There IS no current scope to cut — this skill prunes, it doesn't generate; run aha-mapper first if no loop is defined

The decision rule (from Ch 3/4)

For every item in the current required-onboarding list, ask:

If I removed this, would the user still hit the aha moment?

Yes (they'd still hit it): the item is Helpful, not Required. Move it out of the activation flow. Surface it later via progressive disclosure — empty-state nudge, contextual tooltip, dismissible banner after the first aha.
No (removing it blocks the aha): the item is Required. Keep in onboarding, but verify it's the minimum version (one calendar not five, one event type not multiple).

The book's stronger heuristic: cut the required list in half. Most PMs over-estimate what's required because every team owner argues their thing belongs in the flow. Forcing a 50% cut surfaces what is actually load-bearing.

Progressive disclosure for Helpful items

Each item moved to Helpful needs a re-introduction plan, not just deletion:

First-aha trigger — nudge appears right after the user completes the smallest loop once
Contextual surface — the nudge fires at the moment the user would benefit (when they create their second event type, prompt for default settings)
Dismissible — user can skip and the nudge doesn't reappear for that session
Maximum 3 nudges per session — beyond this is harassment

Items the user never taps after 5 sessions are candidates for the "Surface only on user request" bucket (a settings menu, not a proactive prompt).

Refusal mode

If the user insists on keeping the full required list ("but every step is important"), refuse to recommend the unchanged list and quantify the cost: each additional required step costs roughly 5–15% of activations in the book's running numbers. State the trade explicitly.

If items are being moved to Helpful but the user has no nudge plan, flag this — moved-and-buried is not progressive disclosure, it's hiding. Recommend the 4-element plan above.

Inputs

Current required-onboarding step list (or feature list to allocate)
The defined aha moment + smallest loop (output of aha-mapper if available)
Current activation funnel data, if any
(Optional) feature owners who would object — the book's gotcha is internal politics

Output structure

## Scope cut verdict: [HALVED | NEEDS_DEEPER_CUTS | NEEDS_MORE_EVIDENCE]

### Required (blocks the aha)
| Step | Why required | Minimum version |
|---|---|---|
| ... | ... | ... |

### Helpful (cut from onboarding, surfaced later)
| Step | Why cut | Re-surface trigger | Dismissible? |
|---|---|---|---|
| ... | ... | ... | ... |

### The cut
- Original required count: N
- Cut required count: M
- Reduction: ((N-M)/N)*100%

### Nudge sequence
[The 3-step post-aha surfacing plan for Helpful items]

### Expected activation impact
[1-2 sentences citing the book's 5-15%-per-step heuristic]

### Recommendation
[Concrete next step]

### Analogous case from the book
[A named example]

Worked example from the book

Calendly's original onboarding (Ch 4) → cut from 8 to 3 required

Original required steps (the team had argued each was essential):

Sign up
Connect a calendar
Set default timezone
Configure default duration
Add a profile photo
Set buffer time between meetings
Create your first event type
Copy your share link

Cut to required:

Sign up
Connect a calendar
Copy your share link (a default event type is auto-created)

Moved to Helpful, surfaced after first booking:

Default timezone (auto-detected from browser; nudge only if booking came in from a different zone)
Default duration (nudged when user creates a second event type)
Profile photo (banner in user dashboard, dismissible)
Buffer time (advanced settings menu, not proactive)
Multiple event types (nudged after first 3 bookings completed)

Result: cut from 8 required steps to 3. Activation rate roughly doubled. The 5 moved items had no measurable activation impact when removed.

Gotchas

"Required" is a political negotiation, not a fact. Every team feels their thing must be onboarded. Forcing a 50% cut surfaces who has data and who has feelings.
Moving to Helpful without a re-surface plan = burying the feature. If you cut it and never bring it back, you've quietly killed it. Use the 4-element nudge plan or accept the kill.
The "minimum version" of a required step matters. "Connect a calendar" is required, but "connect ALL calendars in your org" is not. Reduce each required step to its smallest viable version.
Auto-detection beats prompting for any field that can be inferred. Timezone, locale, name — pull from browser/account, don't ask.

Anti-patterns to flag

If the user's plan has any of these, surface them:

Required-step list growing past 5 items — the book's running threshold; push back hard
Required fields that have a sensible default (timezone, duration) — default and let users change later, don't block
Moving items to "Optional in settings" without a proactive nudge plan — that's hiding, not disclosing
Required onboarding step that doesn't appear in the smallest loop — by definition it's not required; the team has scope creep

Source

Ch 3 (SLC scope cuts) and Ch 4 (Activation and the smallest loop).

About this skill

This skill is part of the ProductMind Skills catalog — open-source product-judgment skills for AI coding agents.

Source: Building Rocketships by Oji and Ezinne Udezue — Chapter 3. This skill applies one named framework from the book. It works standalone in any agent that supports the SKILL.md standard.

Want the orchestrated version? The RocketshipsGPT agent runs all 13 book skills together, with persistent memory, framework-fidelity evals, voice-checked output, and the book's anti-pattern catches built in. → rocketshipsgpt.productmind.co

License: Skill content licensed under CC BY-SA 4.0. Attribution required; derivative works must share alike. See ADR 0005 for rationale.

Skill version: 1.0.0 · Published: 2026-06-05