The 5-Step Algorithm Skill for Claude

🌐 Language / Sprache: English | Deutsch

A Claude Skill that implements systematic process and requirements analysis using the 5-step algorithm (inspired by Elon Musk's engineering methodology), adapted for public administration, education, and organizational development.

Overview

This skill guides Claude through a rigorously sequential analysis methodology that examines processes, requirements, and organizational structures for necessity, simplicity, and efficiency. Originally developed for hardware engineering at SpaceX/Tesla, the methodology has been carefully adapted for knowledge work in the public sector — with added guardrails for legal compliance, stakeholder involvement, and democratic accountability.

The 5 Steps

The sequence is non-negotiable. Each step must be completed before moving to the next.

1. Question the Requirements — Every requirement is guilty until proven innocent. Identify who created it, what problem it solves, and whether that problem still exists.

2. Delete — Remove everything not classified as valid. Combat addition bias. Conduct mandatory stakeholder analysis before every deletion.

3. Simplify and Optimize — Only optimize what remains. Apply first-principles thinking. Eliminate media breaks, shorten approval cascades, consolidate roles.

4. Accelerate — Increase speed only after simplification. Find bottlenecks, parallelize steps, shorten feedback loops.

5. Automate — Automate last, never first. Only automate stable, predictable, rule-based processes.

Key Adaptations for Public Administration

Unlike the original engineering context, this adaptation includes:

• Chesterton's Fence Test — Understand why a rule exists before removing it
• Mandatory Stakeholder Analysis — Before any deletion decision
• Data Quality Indicators — 🔵 Verified vs. ⚪ Estimate for executive decisions
• Legal Framework Awareness — Distinguish legal obligations from internal habits
• Reversibility Planning — Document deletions for quick reinstatement
• Pilot Projects — Test before organization-wide rollout

Use Cases

This skill activates when users:

• Want to optimize or challenge processes
• Need to evaluate requirements or specification documents
• Want to reduce bureaucracy or streamline workflows
• Need to assess IT systems, tools, or platforms for necessity
• Ask about "simplification," "efficiency," "process optimization," or "lean"
• Want to challenge a project plan or organizational structure
• Mention the "Musk Algorithm" or "5-step method"

Installation

For Claude.ai Projects (Recommended)

1. Download the SKILL.md file from this repository
2. Go to your Claude.ai project → Project Settings → Skills
3. Upload the SKILL.md file or create a new custom skill
4. The skill will automatically activate based on user interaction patterns

For Claude Desktop / API

Include the skill content in your system prompts or custom instructions.

For detailed installation instructions (including API and MCP Server setup), see the Installation Guide.

Example: School Enrollment Process

The repository includes a complete worked example analyzing a school enrollment process that was reduced from 12 steps to 5, cutting processing time from 6 weeks to under 24 hours.

Key results from the example:

• Time savings (admin staff): ~80%
• Time savings (parents): 6 weeks → 24 hours
• Error reduction: ~90%
• Cost reduction: ~CHF 12,000/year

Output Format

Every analysis produces:

• Classified requirements list (🟢 Valid | 🟡 Questionable | 🔴 Deletion candidate)
• Stakeholder analysis with reversibility plan
• Before-and-after process comparison
• Acceleration levers and optimized timeline
• Prioritized automation plan
• Quick-Win Matrix (2×2: Effort vs. Impact)

Theoretical Foundation

The skill integrates concepts from:

• Lean Manufacturing (Toyota Production System) — Waste elimination
• First Principles Thinking — Question conventions, return to fundamentals
• Theory of Constraints (Goldratt) — Identify and optimize bottlenecks
• Chesterton's Fence — Understand before removing
• Addition Bias Research (Nature, 2021) — Humans prefer adding over removing

For the full theoretical background, see Background.

Repository Structure

musk-algorithm-skill/
├── SKILL.md                    # The skill (English)
├── SKILL.de.md                 # The skill (German)
├── README.md                   # Documentation (English)
├── README.de.md                # Documentation (German)
├── LICENSE                     # MIT License
├── CONTRIBUTING.md             # Contribution guidelines (English)
├── CONTRIBUTING.de.md          # Contribution guidelines (German)
├── CHANGELOG.md                # Change log
├── docs/
│   ├── installation.md         # Installation guide (English)
│   ├── installation.de.md      # Installation guide (German)
│   ├── background.md           # Theoretical background (English)
│   └── background.de.md        # Theoretical background (German)
└── examples/
    ├── example-1-school-enrollment.md    # Worked example (English)
    └── example-1-schulanmeldungen.de.md  # Worked example (German)

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

Areas for potential improvements:

• Additional domain-specific adaptations (healthcare, finance, NGOs)
• Integration examples with specific platforms
• Evaluation metrics for process optimization quality
• Additional worked examples from different sectors

License

MIT License — see LICENSE file for details.

Author

Hayal Oezkan
Marketing and Communication / AI Specialist Group, City of Zurich Administration

GitHub: @malkreide

Acknowledgments

Based on Elon Musk's 5-Step Algorithm as documented in Walter Isaacson's biography Elon Musk (2023), with adaptations drawing from:

• Lean Manufacturing (Toyota Production System)
• Theory of Constraints (Eliyahu Goldratt)
• Chesterton's Fence (G.K. Chesterton)
• Addition Bias research (Adams et al., Nature 2021)

---

"If you're not occasionally adding things back in, you're not deleting enough." — Elon Musk

---

Made with ❤️ in Zurich

LinkedIn • Documentation • Contributing