Claude Code Plugins

Community-maintained marketplace

Feedback

roadmap-builder

@fbeach73/rpl
0
0

Helps prioritize features and decide what to build next. Use when advising on roadmap, challenging feature ideas, evaluating what to build, or keeping product focused. Applies Impact vs Effort framework and stage-based rules to prevent feature creep.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name roadmap-builder
description Helps prioritize features and decide what to build next. Use when advising on roadmap, challenging feature ideas, evaluating what to build, or keeping product focused. Applies Impact vs Effort framework and stage-based rules to prevent feature creep.

Roadmap Builder

Prioritize features and keep roadmap focused on what matters.

Prioritization Framework

Impact vs Effort Matrix:

  1. High Impact, Low Effort → Build first
  2. High Impact, High Effort → Consider after quick wins
  3. Low Impact, Low Effort → Maybe, if time permits
  4. Low Impact, High Effort → Don't build

Impact = How much it moves the needle on:

  • User retention
  • Core use case improvement
  • Revenue
  • User acquisition

Effort = Time to build + maintain:

  • <1 week = Low
  • 1-2 weeks = Medium
  • 2 weeks = High

Category Priority Order

Build in this order:

  1. Retention - Keep users coming back
  2. Core Features - Make core use case better
  3. Monetization - Convert users to paying
  4. Growth - Get new users

Never work on Growth before Retention is solid.

Stage-Based Rules

Pre-Launch (No users yet)

ONLY build:

  • Core loop features
  • Minimum features for user to complete main use case

DO NOT build:

  • Nice-to-haves
  • Advanced features
  • Admin dashboards
  • Analytics
  • Auth (unless core to the product)
  • Any feature "we might need later"

Goal: Ship core loop, get first user, validate idea

Post-Launch (0-100 users)

ONLY build:

  • Features users explicitly request
  • Fixes for things that block users

DO NOT build:

  • Features you think users want
  • Optimizations
  • Scale features
  • Features "power users might want"

How to decide:

  • Did 3+ users ask for this?
  • Is it blocking users from core loop?
  • If neither → don't build

Goal: Find product-market fit

Growth Phase (>100 active users)

Prioritize:

  • Features that reduce churn
  • Features that increase sharing/referrals
  • Monetization features (if not done)

Validate first:

  • Can we fake it to test demand?
  • What's the usage data saying?
  • Is this asked for or just nice?

Goal: Scale what's working

Feature Validation Questions

Ask these three questions for EVERY feature:

  1. Does this serve the core use case?

    • If no → don't build
    • If "kind of" → probably don't build
  2. Will users actually use this or just say they want it?

    • Check: Are they asking because they hit this problem?
    • Or: Is it a "would be nice" idea?
    • If unclear → fake it first
  3. Can we fake it first to validate demand?

    • Manual process
    • No-code tool
    • Hardcoded version
    • If yes → fake it before building

Red Flags

Feature Creep:

  • "Since we're building X, we should also add Y"
  • "This would be cool to have"
  • "Competitors have this"
  • Building features not explicitly requested

Premature Optimization:

  • "We need to handle scale"
  • "Let's build this properly from the start"
  • "We should future-proof this"
  • Before you have users at scale

Imaginary Users:

  • "Power users will want..."
  • "Enterprise customers need..."
  • "People might use this for..."
  • Before real users ask for it

Any of these? Stop and re-evaluate.

Decision Tree

When asked "Should we build X?":

  1. What stage are you in?

    • Pre-launch → Is it core loop? No → Don't build
    • Post-launch → Did users ask? No → Don't build
    • Growth → Does it reduce churn? No → Evaluate further
  2. If users asked:

    • How many users? <3 → Don't build yet
    • Is it blocking them? No → Lower priority
    • Can you fake it? Yes → Fake it first
  3. Score Impact & Effort:

    • High impact, low effort → Build
    • High impact, high effort → Evaluate ROI
    • Low impact, any effort → Don't build
  4. Which category?

    • Retention → High priority
    • Core → Medium priority
    • Monetization → If retention is solid
    • Growth → If retention + core are solid

Default answer: "Don't build it yet."

Roadmap Output Format

When creating roadmap:

Now (This week):

  • Feature: High impact, low effort, users asking
  • Feature: Blocks core loop

Next (This month):

  • Feature: High impact, medium effort, validated demand
  • Feature: Reduce churn for existing users

Maybe (Future):

  • Feature: Good idea but low priority
  • Feature: Wait for more user requests

Not Building:

Challenging Feature Ideas

When user proposes a feature, respond with:

  1. Validation questions:

    • Who asked for this?
    • What problem does it solve?
    • Can we fake it first?
  2. Priority assessment:

    • Stage-appropriate?
    • Impact vs effort?
    • Category priority?
  3. Alternative approaches:

    • Manual workaround?
    • Existing tool integration?
    • Simpler version?
  4. Recommendation:

    • Build now / Build later / Don't build
    • Reasoning in 1-2 sentences

Be direct. Better to push back now than build wrong thing.

Common Mistakes

❌ Building features because competitors have them
❌ Adding features before core loop works
❌ Building for hypothetical future users
❌ Optimizing before you have users
❌ Saying yes to every request
❌ Building what you find interesting vs what users need
❌ Working on growth before retention

Quick Checklist

Before adding to roadmap:

  • Is this explicitly requested by users?
  • Does it serve core use case?
  • Can't we fake it first?
  • Is it appropriate for our stage?
  • High impact, reasonable effort?
  • Right priority category?
  • Not a red flag scenario?