A lightweight, cross-platform desktop app for building a modern AI skills manager.
Skill Studio is a Tauri-based desktop foundation for managing AI skills with a cleaner architecture, a better UX baseline, and a stronger separation between installation, activation, backup, sync, and diagnostics.
It is designed as an improved standalone successor to the skills management experience explored from the cc-switch project.
- Lightweight desktop stack powered by Tauri 2 + Rust
- Modern frontend with React 19, TypeScript, Vite, Tailwind CSS v4, and React Query
- Cross-platform architecture for macOS / Windows / Linux evolution
- Modular backend boundaries for manifest, installer, sync, backup, and diagnostics
- Polished dashboard scaffold with Installed / Sources / Backups / Settings starter pages
- GitHub-ready project structure for iterating toward a production desktop app
Current repository version includes:
- Desktop shell layout
- Hash-router based navigation
- Feature placeholder pages
- Tauri command examples (
ping,get_app_overview) - Buildable frontend and Rust scaffold
This repository is currently a polished scaffold / architecture baseline, not the final fully featured product.
- React 19
- TypeScript
- Vite
- Tailwind CSS v4
- React Router
- React Query
- Lucide React
- Tauri 2
- Rust
- tauri-plugin-log
The goal is not just to recreate an existing skills manager, but to build a better one with clearer domain boundaries.
Planned separation includes:
- Skill Source: repository / ZIP / local import origin
- Skill Manifest: normalized metadata contract
- Installed Skill: workspace-owned installed entity
- Activation / Deployment: per-app enable / disable state and sync strategy
- Backup / Restore: explicit lifecycle support instead of hidden side-effects
- Diagnostics: drift detection, broken links, invalid states, and health reporting
Make sure your environment has:
- Node.js
- npm
- Rust toolchain
- Tauri development prerequisites for your OS
git clone git@github.com:cikichen/skill-studio.git
cd skill-studio
npm install
npm run tauri devnpm run buildcd src-tauri
cargo buildsrc/
app/ App router entry
features/
skills/ Installed skills experience
sources/ Sources & discovery management
backups/ Backup / restore flow
settings/ Application settings
shared/
components/ Layout shell and reusable UI blocks
lib/ Tauri bridge helpers
types/ Frontend domain types
src-tauri/src/
commands/ Tauri invoke command layer
core/
manifest/ Manifest parsing contracts
installer/ Install orchestration skeleton
sync/ Deployment strategy skeleton
backup/ Backup lifecycle skeleton
diagnostics/ Health-check skeleton
state.rs Shared application state
lib.rs App bootstrap and command registration
- Dark polished dashboard shell
- Installed / Sources / Backups / Settings starter pages
- Hash Router based desktop navigation
- React Query provider setup
- Tauri command bridge examples
- Rust module boundaries ready for real implementation
- Project plan document:
PROJECT-PLAN.md
- Define
skill.jsonmanifest schema - Replace mock page data with real Tauri commands
- Add workspace discovery and target app detection
- Implement install / uninstall / enable / disable workflows
- Implement backup / restore / diagnostics flows
- ZIP import workflow
- Repository discovery management
- Unmanaged skill scan
- Conflict detection and source resolution
- Better component system in shadcn/ui style
- Packaging, icons, metadata, and release automation
- Routing currently uses Hash Router for desktop friendliness
- Backend currently exposes example commands only
- The current UI is a starter shell intended for rapid iteration
PROJECT-PLAN.mdcontains the implementation planning baseline
MIT