agentleFS
neoforge-devneoforge-dev/forge/.claude/skills/mvp-spec-writer/SKILL.md

mvp-spec-writer

Skill0 starsChanged 7 months agoMITView on GitHub

Auto-generate features.json from exploration reports with prioritized backlog and complexity estimates

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create features following this priority framework:

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

MVP Spec Writer Skill

Converts exploration reports into actionable features.json files with prioritized backlog, acceptance criteria, and complexity estimates. Bridges the gap between exploration and building.

When to Use

  • After completing /niche-explorer
  • Starting a new MVP build
  • Refining an existing features.json
  • Creating structured backlog from product requirements

Prerequisites

  • Exploration report exists (from /niche-explorer)
  • Domain and project directories exist
  • Understanding of FORGE feature schema

Inputs

exploration: babybit-es/explorations/pedi-sync-exploration.md
project_name: pedi-sync
constraints:
  backend: FastAPI
  auth: JWT
  max_p0_features: 7
  target_completion: 2 weeks

Workflow

Step 1: Parse Exploration Report

Extract key information:

  • MVP name and description
  • Problem statement
  • Target users
  • Differentiator
  • Tech stack requirements
  • Success metrics

Step 2: Generate Feature Backlog

Create features following this priority framework:

P0 (Must Have - MVP Blocker)

  • Core value proposition features
  • Essential user flows
  • Basic auth/security
  • Minimum data model

P1 (Should Have - Launch Enhancer)

  • Enhanced user experience
  • Additional workflows
  • Performance optimizations
  • Analytics/tracking

P2 (Nice to Have - Post-Launch)

  • Advanced features
  • Integrations
  • Admin tools
  • Optimization

Step 3: Structure Each Feature

{
  "id": "PS-001",
  "title": "Project scaffold with FastAPI backend",
  "priority": "P0",
  "type": "backend|frontend|infrastructure|content",
  "complexity": "low|medium|high",
  "estimated_hours": 4,
  "acceptance_criteria": [
    "FastAPI app with health endpoint returns 200",
    "SQLAlchemy async setup with User model",
    "Pytest configuration with 1+ passing test",
    "Docker setup with hot reload"
  ],
  "dependencies": [],
  "technical_notes": "Use fastapi-service-template skill"
}

Step 4: Estimate Complexity

Use these guidelines:

Low Complexity (2-4 hours)

  • CRUD endpoints
  • Simple UI components
  • Standard auth flows
  • Basic data models

Medium Complexity (4-8 hours)

  • Business logic workflows
  • Complex forms
  • File uploads
  • Email notifications

High Complexity (8-16 hours)

  • Third-party integrations
  • Real-time features
  • Complex algorithms
  • Multi-step workflows

Step 5: Build Dependency Graph

Identify feature dependencies:

  • Infrastructure before backend
  • Backend models before API endpoints
  • API endpoints before frontend
  • Auth before protected routes

Step 6: Generate features.json

Output complete features.json:

{
  "project": "pedi-sync",
  "domain": "babybit-es",
  "version": "1.0.0",
  "description": "Share baby nutrition data with pediatricians",
  "tech_stack": {
    "backend": "FastAPI",
    "frontend": "React PWA",
    "database": "PostgreSQL",
    "auth": "JWT",
    "deployment": "Railway + Cloudflare Pages"
  },
  "compliance": ["COPPA"],
  "success_metrics": [
    "100 active parents in 30 days",
    "70% user retention after 7 days",
    "4.5+ app store rating"
  ],
  "features": [
    {
      "id": "PS-001",
      "title": "Project scaffold with FastAPI backend",
      "priority": "P0",
      "type": "infrastructure",
      "complexity": "medium",
      "estimated_hours": 4,
      "acceptance_criteria": [
        "FastAPI app with health endpoint",
        "SQLAlchemy async setup",
        "Pytest configuration",
        "Docker setup"
      ],
      "dependencies": [],
      "technical_notes": "Use /fastapi-service-template skill"
    },
    {
      "id": "PS-002",
      "title": "User authentication with JWT",
      "priority": "P0",
      "type": "backend",
      "complexity": "medium",
      "estimated_hours": 6,
      "acceptance_criteria": [
        "Parent account registration",
        "Login with email/password",
        "JWT token generation",
        "Protected endpoints",
        "Password reset flow"
      ],
      "dependencies": ["PS-001"],
      "technical_notes": "Parents are account owners (COPPA compliance)"
    },
    {
      "id": "PS-003",
      "title": "Feeding log data model",
      "priority": "P0",
      "type": "backend",
      "complexity": "medium",
      "estimated_hours": 5,
      "acceptance_criteria": [
        "FeedingLog model with timestamp, food, quantity",
        "Baby profile model",
        "CRUD endpoints for feeding logs",
        "Filtering by date range",
        "JSON export endpoint"
      ],
      "dependencies": ["PS-001", "PS-002"],
      "technical_notes": "Support multiple babies per parent account"
    },
    {
      "id": "PS-004",
      "title": "Feeding log entry form",
      "priority": "P0",
      "type": "frontend",
      "complexity": "medium",
      "estimated_hours": 6,
      "acceptance_criteria": [
        "Quick-add feeding log form",
        "Food search/autocomplete",
        "Quantity input (oz, ml, servings)",
        "Timestamp picker (defaults to now)",
        "Offline support"
      ],
      "dependencies": ["PS-003"],
      "technical_notes": "PWA offline-first for parents on the go"
    },
    {
      "id": "PS-005",
      "title": "Export feeding history for pediatrician",
      "priority": "P0",
      "type": "backend",
      "complexity": "high",
      "estimated_hours": 8,
      "acceptance_criteria": [
        "Generate PDF report with feeding summary",
        "Include growth charts (if available)",
        "Filter by date range (default: last 30 days)",
        "Email or download options",
        "Professional medical format"
      ],
      "dependencies": ["PS-003"],
      "technical_notes": "Use weasyprint for PDF generation"
    },
    {
      "id": "PS-006",
      "title": "Landing page with product demo",
      "priority": "P0",
      "type": "frontend",
      "complexity": "medium",
      "estimated_hours": 5,
      "acceptance_criteria": [
        "Hero section with value proposition",
        "Feature highlights (3-5 sections)",
        "Screenshot carousel",
        "CTA buttons (Download app)",
        "Responsive design"
      ],
      "dependencies": [],
      "technical_notes": "Use /frontend-design skill for distinctive UI"
    },
    {
      "id": "PS-007",
      "title": "Content library (blog posts, social)",
      "priority": "P1",
      "type": "content",
      "complexity": "low",
      "estimated_hours": 3,
      "acceptance_criteria": [
        "5 blog post outlines",
        "10 social media posts",
        "Email templates (welcome, reminder)",
        "SEO-optimized titles",
        "Spanish language versions"
      ],
      "dependencies": [],
      "technical_notes": "Use /content-library-producer skill"
    }
  ],
  "estimated_total_hours": 37,
  "estimated_timeline": "2 weeks",
  "p0_feature_count": 6,
  "p1_feature_count": 1,
  "p2_feature_count": 0
}

Quality Criteria

A good features.json includes:

  • 5-7 P0 features (MVP core)
  • Clear acceptance criteria for each feature
  • Realistic complexity estimates
  • Proper dependency ordering
  • Technical notes for guidance
  • Compliance requirements identified
  • Success metrics defined
  • Total effort estimate
  • All required metadata

Feature Schema Reference

interface Feature {
  id: string;              // Format: {PROJECT_PREFIX}-{NUMBER}
  title: string;           // Short, actionable title
  priority: "P0" | "P1" | "P2";
  type: "backend" | "frontend" | "infrastructure" | "content";
  complexity: "low" | "medium" | "high";
  estimated_hours: number;
  acceptance_criteria: string[];
  dependencies: string[];  // IDs of prerequisite features
  technical_notes?: string;
}

Example Invocation

/mvp-spec-writer

Exploration: babybit-es/explorations/pedi-sync-exploration.md
Project: pedi-sync
Constraints:
  - Backend: FastAPI
  - Auth: JWT with parent accounts
  - Max P0 features: 7
  - Target: 2 weeks completion
  - Must be COPPA compliant

Output

Creates {domain}/{project}/features.json ready for Ralph Loop consumption.

Related Skills

  • /niche-explorer - Generate exploration report (prerequisite)
  • forge loop run -d <domain> -p <project> - Execute features.json
  • /living-docs update - Document the new project

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.