System Design Doc
ragrahari/claude-skills
Gera docs estruturados de design de sistema com arquitetura e análise
Ver no GitHubSkill que gera documentos estruturados de design de sistema a partir de requisitos e contexto. Produz análise completa incluindo problema, arquitetura, modelo de dados, APIs, trade-offs e pontos críticos — ideal para entrevistas técnicas ou planejamento de projetos.
Instalação
npx convez-ai ragrahari/claude-skills1 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
- Preparação para entrevistas de system design
- Documentação técnica de novos projetos de software
- Análise de arquitetura e trade-offs de sistemas complexos
Veredito

Skill bem estruturada que automatiza a geração de design docs em nível de entrevista ou produção. Útil para engenheiros que precisam documentar sistemas rapidamente sem repetir preferências de formato — funciona por trigger automático quando detecta intenção de design.
README
A tradução para o português deste README ainda está na fila.
claude-skills
A growing library of Claude skill files for software engineers — covering system design, code review, documentation, and more.
Skill files teach Claude environment-specific best practices so you get consistently structured, opinionated output without re-explaining your preferences every time.
What's a skill file?
A skill file (SKILL.md) is a markdown file you install into Claude. When you ask something that matches the skill's trigger, Claude reads the skill first and follows its instructions — producing output that's structured, thorough, and tailored for engineers.
Think of it as a reusable "how I want Claude to work" config that travels with you across projects and machines.
Skills
| Skill | What it does | Trigger phrases |
|---|---|---|
system-design-doc | Generates a full system design document — problem statement, architecture diagram, data model, API design, trade-offs, and gotchas | "design a X system", "create a design doc for Y", "how would you architect Z" |
More skills coming. Contributions welcome — see CONTRIBUTING.md.
Installation
Claude.ai (web / desktop / mobile)
- Download the
.skillfile from the releases page or the skill's folder - Open Claude.ai → Settings → Skills → Install from file
- Select the
.skillfile - Done — Claude will now use the skill automatically when relevant
Claude Code (CLI)
claude skill install system-design-doc.skill
Usage
Once installed, just describe what you want — no special syntax needed.
Examples that trigger system-design-doc:
Design a URL shortener like bit.ly
How would you architect a real-time notification service for 10M users?
Create a design doc for a ride-sharing app backend
Help me think through the design for a distributed job scheduler
Claude will ask a few clarifying questions if context is missing, then produce a structured doc covering:
- Problem Statement
- Requirements (functional + NFRs)
- Capacity Estimation
- High-Level Architecture (with diagram, if you have connected tools)
- Data Modeling
- API Design
- Scalability & NFR Mapping
- Trade-off Analysis
- Gotchas & Failure Modes
- Out of Scope / Future Considerations
Output format: Ask for Markdown (default) or a Word doc (.docx) you can share with your team.
Depth: Default is interview-level (concise, readable in 10 minutes). Say "production-level" or "for my team" for a more detailed output.
Example output
See skills/system-design-doc/examples/url-shortener.md for a full sample output.
Building the .skill file yourself
If you've cloned the repo and want to package a skill locally:
# Requires the Anthropic skill creator scripts
# Clone this repo, then:
cd claude-skills
python -m scripts.package_skill skills/system-design-doc ./dist
Or just download the pre-packaged .skill file from releases.
Contributing
Contributions are welcome — new skills, improvements to existing ones, or better example outputs.
See CONTRIBUTING.md for guidelines. The short version:
- Open an issue to propose a new skill before building it
- Each skill needs a
SKILL.mdand at least one example output - PRs should include a brief description of what the skill does and 2–3 example trigger phrases
Roadmap
Skills planned or in progress:
java-code-review— structured Java code review with thread safety, exception handling, and complexity checksunit-test-generator— JUnit/Mockito test generation with edge cases and mocking strategypr-description-writer— structured PR descriptions from a diff or change summaryadr-writer— Architecture Decision Record generator
Have an idea? Open an issue.
License
MIT — free to use, share, and build on.