Claude Code OS — Platypus AI
rafabellan/ccos-en
Sistema inteligente que configura seu workspace de IA em 5 minutos
Ver no GitHubUm starter kit que entrevista você sobre seu negócio, aprende seu contexto e gera um workspace personalizado com skills prontos para seu trabalho diário. Inclui comandos slash para propostas de vendas, análise de dados, criação de conteúdo, gestão de conhecimento e publicação na web.
Instalação
npx convez-ai rafabellan/ccos-en1 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
- Criar propostas de vendas e emails profissionais automaticamente
- Analisar dados e gerar resumos executivos com insights
- Construir e consultar um vault de conhecimento pessoal
- Gerar conteúdo para redes sociais (carrosséis, scripts, slides)
- Organizar contexto de negócio e projetos em um sistema estruturado
Veredito

Kit bem estruturado e prático que reduz tempo de setup inicial do Claude Code. Útil para profissionais e pequenos negócios que querem um workspace organizado, mas demanda JavaScript e familiaridade com Claude Code para melhor aproveitamento.
README
A tradução para o português deste README ainda está na fila.
Claude Code OS — Platypus AI
A starter kit for using Claude Code in your daily work. The system interviews you, learns your business, generates your workspace, and ships with skills to handle the work you actually do.
How to install
Option 1 — Via prompt (easiest)
With Claude Code open in any folder, paste this prompt:
Install the repo https://github.com/rafabellan/ccos-en.git into the current folder, open it, and run /setup
Claude does the rest: clones the repo, enters the folder, and starts setup.
Option 2 — Via terminal
1. Clone the repo
git clone https://github.com/rafabellan/ccos-en.git
cd ccos-en
2. Open in VS Code
code .
3. Open the integrated terminal (Ctrl + on Windows / Cmd + on Mac) and run:
claude
4. Call setup
/setup
Claude will ask you a few questions and configure the system for your business. About 5 minutes and you're set.
What's in the kit
Slash commands ready to use:
/setup— configures the system for your business (start here)/start— loads business context at the beginning of each work session/sync— saves your work to GitHub (commit + push, configures on first run)/carousel— creates carousels for Instagram and TikTok using your visual identity/sales-proposal— generates a professional HTML proposal from a brief/publish-site— publishes any HTML on the web with a shareable link/slide— creates a visual slide/card for presentations/analyze-data— analyzes a file and generates an executive summary with insights/post-script— turns an idea or text into a post or video script/pro-email— drafts a professional email from free-form context/ingest— adds a source (book, course, video, article, notes) to your knowledge vault/query— asks a question against your knowledge vault, with citations/lint— health-checks the knowledge vault (orphans, missing links, thin pages)/refresh— sweeps the project and updates any context files that drifted/new-project— creates a new project folder with its own dedicated CLAUDE.md (interviews you about the project)
Folders generated by /setup:
_context/— your business context and preferencesbrand/— visual identity guide for your brandknowledge/— your personal Obsidian-ready knowledge vault, with domains set up from your setup interview answerstemplates/tools/catalog.md— APIs, CLIs, and MCPs available to use inside skills
data/ folder:
- Drop zone for files you want to analyze (CSV, XLSX, TXT, PDF)
- Useful when you don't have a Google Drive MCP installed
- Use with
/analyze-data data/your-file.csv
data/ vs knowledge/raw/: data/ is for one-off files you want to analyse right now and probably won't look at again (a client's CSV export, a quarterly report PDF, a survey dump). knowledge/raw/ is for sources you want to keep, cross-reference, and build on over months. If you're going to ingest it and let it compound, it belongs in knowledge/raw/. If you just want a summary and chart, drop it in data/.
Stuck?
For any question or help, bring it to the community. That's where you get support and talk to me directly.
To join the Platypus Lab community, visit skool.com/platypus-lab.