The dashboard for your AI coding assistants
A free desktop app for managing AI coding assistant context files. Scan your projects, edit context files, and track token counts across Claude Code, Cursor, GitHub Copilot, Windsurf, and more.
If you use AI coding assistants, you know the pain:
- Context chaos: CLAUDE.md here, .cursorrules there, copilot-instructions.md somewhere else
- Token mystery: Is your context file too long? Too short? You have no idea
- Project sprawl: Dozens of projects, each with different AI configurations
- Tool switching: Different AI tools need different file formats
TokenCentric solves all of this with a single, beautiful desktop app.
Manage context files for all major AI coding assistants in one place:
| Tool | File Pattern |
|---|---|
| Claude Code | CLAUDE.md |
| Cursor | .cursorrules |
| GitHub Copilot | .github/copilot-instructions.md |
| Windsurf | .windsurfrules |
| ChatGPT/OpenAI | AGENTS.md |
Real-time, accurate token counts using official tokenizers:
- Anthropic tokenizer for Claude files
- tiktoken (OpenAI) for Cursor, Copilot, Windsurf, and OpenAI files
Color-coded indicators show when files are getting too large.
Full-featured code editor with:
- Syntax highlighting
- Markdown preview (side-by-side or toggle)
- Keyboard shortcuts (Cmd+S to save)
- Unsaved changes indicator
- Scan directories to find all context files
- Create new files from 7 built-in templates
- Edit, duplicate, and delete files
- Reveal in Finder/Explorer for quick access
Start new projects fast with pre-built templates:
- Minimal
- Full-Stack JS (Node, React, TypeScript)
- Python
- Monorepo
- API Project
- Mobile (React Native/Expo)
- Careful AI (safety-focused rules)
Automatic theme detection with manual override. Looks great day and night.
Always stay up to date with automatic updates via GitHub Releases.
Download the latest release for your platform:
| Platform | Download |
|---|---|
| macOS | Download .dmg |
| Windows | Download .exe |
# Clone the repository
git clone https://github.com/helsky-labs/tokencentric.git
cd tokencentric
# Install dependencies
npm install
# Build for your platform
npm run dist# Install dependencies
npm install
# Run in development mode
npm run dev
# In a separate terminal, start Electron
npm start| Script | Description |
|---|---|
npm run dev |
Start all dev watchers (main, preload, renderer) |
npm start |
Launch Electron app |
npm run build |
Build all TypeScript and Vite |
npm run dist |
Build and package for current platform |
npm run dist:mac |
Build and package for macOS |
npm run dist:win |
Build and package for Windows |
npm run lint |
Run ESLint |
npm run format |
Format code with Prettier |
- Framework: Electron 28
- Frontend: React 18 + TypeScript + Vite
- Styling: Tailwind CSS
- Editor: Monaco Editor
- Tokenizers: @anthropic-ai/tokenizer, tiktoken
- Storage: electron-store
- Updates: electron-updater
TokenCentric is free and open source. If you find it useful, consider:
- Starring the repo
- Supporting via PIX (Brazilian instant payment):
772337c9-12fc-47fa-8849-32fb5f696129 - Sharing with friends who use AI coding assistants
- File watching (auto-refresh on external changes)
- Combined token count for inheritance chains
- Search/filter files
- Custom template creation
- Custom tool profile creation
- Template import/export
- "What context does AI see here?" visualization
MIT
TokenCentric is a Helsky Labs project. Check out our other tools for developers and creators.