S5-ProjectManager: Analisador de Projetos com Framework Bill Gates
yqing999-CmYk/ProductCriticAgent
Análise estratégica de projetos usando framework Bill Gates em 5 dimensões
Ver no GitHubSkill que aplica o framework de tomada de decisão estratégica de Bill Gates para avaliar se projetos merecem investimento de tempo e recursos. Analisa projetos através de 5 dimensões independentes (viabilidade tecnológica, controle de custos, alocação de recursos, inovação e dinâmica de mercado) e entrega recomendações diretas sobre viabilidade, oportunidades de melhoria e descartes.
Instalação
npx convez-ai yqing999-CmYk/ProductCriticAgent1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Decisões go/no-go de project managers e CTOs
- Avaliação de iniciativas estratégicas por CEOs e founders
- Priorização de roadmap para empreendedores solo
- Análise de saúde de portfólio em times de liderança
Veredito

Ferramenta sólida e pragmática para quem precisa de critério objetivo na aprovação de projetos. O framework das 5 dimensões é bem estruturado e evita viés emocional, mas funciona melhor em contextos onde dados estão disponíveis e métricas de mercado são mensuráveis.
README
A tradução para o português deste README ainda está na fila.
S5-ProjectManager: Bill Gates Project Analyst Skill
Introduction
S5-ProjectManager is a Claude AI skill that mimics Bill Gates' strategic decision-making framework to analyze whether projects deserve time and resource investment. Rather than providing financial ROI predictions, it delivers straight-to-the-point strategic evaluations of project viability, improvement opportunities, and removal recommendations.
Primary Audience
- Project managers and CTOs making go/no-go decisions
- CEOs and founders evaluating strategic initiatives
- Solo entrepreneurs and freelancers prioritizing their roadmap
- Leadership teams assessing portfolio health
Core Philosophy
This skill applies Bill Gates' proven methodology from his Microsoft era and modern philanthropy mindset—combining systems thinking, data-driven analysis, cost awareness, and long-horizon vision to cut through noise and focus on what matters.
How It Works
The 5-View Analysis Framework
The skill evaluates projects across five independent dimensions, each scored 1-10:
1. Technology Doable
- Can this be implemented and scaled reliably?
- Evaluates strength points, weak points, pain points, implementation difficulty
- 1 = Impossible; 10 = Proven & scalable
- Red flag: If cost/difficulty exceeds 10x industry baseline → Bad project
2. Cost Big Cutter
- Can costs be controlled while maintaining quality?
- Evaluates total implementation cost vs. industry average
- 1 = Cost prohibitive; 10 = Efficient & cost-controlled
- Rule: If cost > 10x average → Reject immediately
3. Resource Allocation Pusher
- Is this the best quality-to-cost ratio available?
- Combines cost awareness with quality expectations
- 1 = Poor allocation; 10 = Optimal allocation
- Green flag: Cost ≤ 2x average + top-tier quality
4. Inventor (Innovation)
- Is this fundamentally new or a copy of existing solutions?
- Evaluates novelty of technology or business model
- 1 = Copy of existing; 10 = Entirely novel market creation
- Note: Innovation alone doesn't guarantee success—must combine with other views
5. Part of a Tidal Wave
- Is there significant market momentum or adoption rate?
- Evaluates consumer behavior shifts and industry dynamics
- 1 = No movement; 10 = Massive shift
- Green flag: Strong tidal wave indicates high potential for rapid scaling
Project Classification
Projects fall into four types, each with unique evaluation priorities:
| Type | Definition | Key Evaluation |
|---|---|---|
| I: Application/Software/Web | Standalone tools, SaaS, web apps | User acquisition, defensibility |
| II: Hardware | Physical products, devices | Manufacturing cost, supply chain |
| III: Service | Human-driven or platform services | Scalability, performance, unit economics |
| IV: System | Infrastructure, platform, ecosystem | Compatibility, network effects, monopoly potential |
Best outcome: Project evolves toward Type IV (System) with monopoly defensibility.
Judgment Criteria
A project is "GOOD" if it meets 3+ of these conditions:
- ≥3 views score 6 or above (strong)
- Technology view ≥6 (implementable)
- Cost view ≥6 (controlled)
- Clear articulation of project type with path to system-level scale
A project is "BAD" if:
- Any view scores ≤3 (critical weakness)
- Cost exceeds 10x industry average (fatal flaw)
- No clear type or path to defensibility
Behavioral Rules (Bill Gates' Decision Style)
- Think in systems, not solutions - Reframe every product as infrastructure/platform problem
- Challenge assumptions directly - Deliver bad news immediately without softening
- Cite data or demand it - Opinion without evidence is noise
- Long-horizon thinking - Evaluate on 10-year arc, not quarterly goals
- Acknowledge imperfection - Gates is competitive but reflective and open to being wrong
- Intellectual humility - Genuine openness about past mistakes
- Apply philanthropy logic - For public-good projects: "Is this scalable to hardest-to-reach populations?"
- Evolve with context - Tone shifts from early combative Gates to post-2000 reflective approach
How to Setup Environments
Prerequisites
- Claude AI Access: Claude Code CLI or web interface (claude.ai/code)
- Python 3.10+ (optional, for local testing)
- Web Access: Skill performs research via Claude API tools (web search, fetching)
Installation Steps
-
Clone or download the skill files
cd /path/to/skills cp -r S5_Project_Manager /your/destination -
Verify skill structure Ensure these files exist in the skill directory:
SKILL.md- Core knowledge base and rulesREADME.md- Documentation (this file)- Any additional knowledge files or examples
-
No external dependencies required
- Skill uses Claude's built-in tools (web search, data analysis)
- No additional Python packages or database setup needed
-
Configuration (Optional)
- Customize knowledge base examples in
SKILL.mdif applying to specific industry - Add domain-specific comparable projects for benchmarking
- Customize knowledge base examples in
How to Run It in Future
Via Claude Code CLI
claude ask-skill S5-ProjectManager "Bill, we have a startup that uses AI to diagnose malaria from a phone camera. Should we fund it?"
Via Claude Web Interface (claude.ai/code)
- Open Claude Code at https://claude.ai/code
- Select the S5-ProjectManager skill from your skill library
- Submit your project analysis request in natural language
Example Prompts
Evaluating a new application:
"Analyze this SaaS product: a Figma plugin that auto-generates accessible React components from design mockups. Target market: 10k-100k developers. Current tech stack: TypeScript, WebAssembly. Estimated development cost: $200k. Team: 4 engineers."
Evaluating a business pivot:
"Our e-commerce company currently does B2C dropshipping. We're considering pivoting to B2B wholesale infrastructure. Is this a good use of our capital?"
Evaluating a hardware startup:
"We're building a home energy monitoring device that works offline via Bluetooth. Manufacturing in Vietnam. Initial production cost: $85 per unit, retail price: $299. Target: 50k units in year 1."
Expected Output
The skill returns:
- Scores for each of the 5 views (1-10 with justification)
- Project classification (Type I, II, III, or IV)
- Judgment call - Is it good, bad, or needs pivot?
- Specific recommendations - What to modify, remove, or emphasize
- 10-year vision - Strategic direction for long-term success
- Length: ≤300 words (sharp, actionable analysis)
How to Wrap It for Deployment
Package as a Reusable Claude Skill
Option 1: Claude Skill Registry (Recommended)
-
Prepare the skill directory:
S5_Project_Manager/ ├── SKILL.md (knowledge base) ├── README.md (this file) ├── skill.json (metadata, see below) └── examples/ (optional: example analyses) -
Create
skill.jsonmetadata file:{ "name": "S5-ProjectManager", "version": "1.0.0", "description": "Bill Gates-inspired project analyst using 5-view strategic framework", "author": "Your Name", "requires_web_access": true, "models": ["claude-3-5-sonnet", "claude-opus"], "activation_keywords": ["project analysis", "strategic evaluation", "go/no-go", "Bill Gates"], "estimated_tokens": 2000, "category": "business-strategy" } -
Register with Claude Code:
- Upload skill to your Claude Code workspace via CLI:
claude register-skill ./S5_Project_Manager - Or manually in Claude Code settings under "Custom Skills"
- Upload skill to your Claude Code workspace via CLI:
Option 2: Self-Hosted Deployment
-
Version control the skill:
git init git add SKILL.md README.md skill.json git commit -m "Initial S5-ProjectManager skill release" git push origin main -
Create a distribution package:
tar -czf S5_Project_Manager-v1.0.0.tar.gz \ SKILL.md README.md skill.json examples/ -
Share via:
- GitHub releases
- Cloud storage (Google Drive, Dropbox)
- Dedicated documentation site
Option 3: Integrate into Custom Claude Agent
-
Embed SKILL.md into system prompt:
from anthropic import Anthropic client = Anthropic() system_prompt = open("SKILL.md").read() response = client.messages.create( model="claude-opus", max_tokens=1024, system=system_prompt, messages=[{"role": "user", "content": user_query}] ) -
Add tool definitions if connecting to external APIs:
- Web search tool (built into Claude)
- Market data tools (optional for deeper research)
- Cost benchmarking databases (optional)
Deployment Checklist
- SKILL.md reviewed and finalized
- README.md complete with all sections
- skill.json metadata created
- Examples folder populated with 3-5 sample analyses
- Web search capability tested
- Performance tested (response time < 30 seconds)
- Version tagged in git (v1.0.0)
- Documentation deployed (GitHub Pages, docs site, etc.)
- User access configured (share link, CLI registration, or integration)
- Feedback mechanism set up (GitHub issues, email, feedback form)
Maintenance & Updates
- Monitor usage: Track which types of projects are analyzed most
- Iterate on framework: Refine scoring guidance based on real outcomes
- Update examples: Add new benchmark projects as industry evolves
- Versioning: Use semantic versioning (MAJOR.MINOR.PATCH)
- Changelog: Document updates in
CHANGELOG.md
Support & Resources
- Knowledge Base: https://www.heropedia.org/product/product-critic/bill-gates
- Bill Gates' Annual Letters: Available at gates.com
- Reference Books: The Innovators by Walter Isaacson, Business Adventures by John Brooks
- Community: Share analyses and discuss project decisions in your team's Claude workspace
License
This skill is provided as-is for educational and business analysis purposes. Attribution to Bill Gates' framework appreciated but not required.
Last Updated: 2026-06-12