AI-Powered Operating System for Reusable Expert Skills
SkillOS is a productivity-focused AI platform that transforms expert workflows into reusable AI-powered skills.
Instead of repeatedly writing prompts, users can launch a skill, provide input, optionally upload supporting documents, and receive structured AI-generated outputs.
The platform is designed to support professional, academic, legal, development, and creative workflows through a unified workspace.
Version: v0.6.0
Status: Stable MVP
- Professional AI Workspace UI
- Skill Search
- Category Filters
- Theme Switching
- Active Navigation
- Responsive Tablet Layout
- Reusable Expert Skills
- File Upload Support
- Runtime Processing Console
- Copy Output
- Reset Workspace
- Success Notifications
- Workspace Information Cards
- Persistent Skill History
- Timeline Style View
- Latest-First Sorting
- Loading States
- Empty States
- React
- TypeScript
- Vite
- React Router
- FastAPI
- Python
- Ollama
- Local LLMs
Current:
- JSON Persistence
Planned:
- PostgreSQL
frontend/
└── src/
├── app/
├── pages/
├── services/
├── types/
├── index.css
└── main.tsxbackend/
└── app/
├── main.py
├── routes/
├── services/
├── skills/
├── uploads/
└── history_store.pyGET /skills
POST /skills/run
GET /skills/historyPOST /files/uploadPOST /ai/chatcd backend
uvicorn app.main:app --reloadBackend URL:
http://localhost:8000cd frontend
npm install
npm run devFrontend URL:
http://localhost:5173- Select a Skill
- Enter Instructions
- Optionally Upload a File
- Execute Skill
- Receive AI Output
- Review Execution History
- Local Ollama dependency
- No user authentication
- No user accounts
- Shared history for all users
- JSON-based persistence
- No cloud deployment
- No favorites system
- No recent skills tracking
- No multi-user support
- Favorites
- Recent Skills
- History Management
- Additional Workspace Improvements
- Authentication
- User Accounts
- PostgreSQL Migration
- Production Data Layer
- Public MVP Release
- Hosted Deployment
- Production AI Integration
Krushanu Bhatt
MCA Graduate | AI/ML Enthusiast | Full Stack Developer
Personal Project