Tech Stack & Core Dependencies
Relevant source files
The following files were used as context for generating this wiki page:
The Tuturuuu platform is a complex ecosystem of web applications and services managed as a high-performance monorepo. It is designed as an intelligent, open-source operating system for modern work and life, emphasizing the elimination of digital noise through automation and unified AI integration.
The architecture leverages a modern TypeScript-first approach, utilizing high-performance runtimes like Bun and a robust Backend-as-a-Service (BaaS) infrastructure provided by Supabase. This setup supports a wide range of applications including workspace management, AI chatbots, finance tracking, and meeting management.
Sources: README.md:1-15, GEMINI.md:5-15, conductor/tech-stack.md:3-8
Core Architecture & Infrastructure
The project is structured as a Turborepo monorepo, which optimizes build tasks and dependency management across multiple packages and applications. Bun serves as the primary package manager and runtime for development tasks, while Node.js (v22+) is maintained for broader compatibility.
Base Infrastructure Components
| Category | Component | Description |
|---|---|---|
| Monorepo Management | Turborepo | Orchestrates builds, testing, and caching. |
| Package Manager | Bun (v1.3+) | High-speed installation and execution runtime. |
| Runtime | Node.js (v22+) | Secondary runtime and build environment. |
| Version Control | Git | Managed with Conventional Commits and GitHub Actions. |
Sources: conductor/tech-stack.md:3-8, README.md:144-150, GEMINI.md:33-40
System Flow: Monorepo Development Cycle
The following diagram illustrates the standard development and verification flow within the monorepo structure.
Sources: GEMINI.md:43-55, conductor/workflow.md:10-25
Frontend & Web Technologies
The frontend is built primarily on Next.js 16 using the App Router and Turbopack. The styling architecture utilizes Tailwind CSS (v4) with a focus on dynamic color tokens to avoid hard-coded values.
Frontend Library Stack
- Core Framework: Next.js with React 19/20 features.
- Component Primitives: Radix UI and shadcn-ui for accessible UI components.
- Specialized UI: Mantine for complex data components and Framer Motion for high-fidelity animations.
- State Management:
- Server State: TanStack Query (React Query) is MANDATORY for all client-side data fetching.
- Client State: Jotai and Zustand for global and atomic state management.
Sources: conductor/tech-stack.md:10-22, GEMINI.md:120-130, README.md:130-140
Data Fetching Flow
This sequence diagram shows the mandatory interaction between Client Components and the data layer using TanStack Query.
Backend & Database (BaaS)
Supabase serves as the primary backend, providing a unified layer for database management, authentication, and real-time updates.
Supabase Services Integration
- PostgreSQL: The core database utilizing Row Level Security (RLS) for multi-tenant workspace isolation.
- Supabase Auth (GoTrue): Handles authentication and session management.
- Supabase Storage: Manages file assets, with specific implementations for AI chat resources and workspace drives.
- Realtime (Broadcast): Used for low-latency client-to-client synchronization, particularly in Kanban boards, using Supabase Broadcast instead of heavy WAL-based change tracking.
Sources: conductor/tech-stack.md:25-33, GEMINI.md:180-195, packages/ai/src/chat/google/route.ts:40-60
AI & Multi-Model Integration
The platform features a sophisticated AI integration layer that supports multiple providers through a unified gateway. This system is centered around the Vercel AI SDK.
Supported AI Providers
| Provider | Implementation Path | Model Examples |
|---|---|---|
google-vertex/edge | Gemini 2.5 Flash Lite | |
| OpenAI | openai/route.ts | GPT-4o, GPT-3.5 |
| Anthropic | anthropic/route.ts | Claude 3.5 Sonnet |
AI Request Pipeline
The platform processes AI requests by validating user credits, filtering message lengths, and applying system instructions before streaming responses.
Development & Quality Assurance
The project enforces strict quality gates through automated tools and specific coding standards.
- Testing: Vitest for unit and integration testing. A minimum of 80% code coverage is targeted for all modules.
- Linting & Formatting: Biome is used for unified linting and formatting.
- Type Safety: Strict TypeScript usage with automated type generation from the Supabase schema (
bun sb:typegen). - Mobile Verification: For the Flutter-based mobile app,
bun check:mobileexecutes Dart formatting, analysis, and testing.
Sources: GEMINI.md:45-55, conductor/tech-stack.md:46-52, conductor/workflow.md:40-60
Conclusion
The Tuturuuu tech stack is designed for extreme scalability and developer velocity. By combining the speed of Bun and Turborepo with the flexibility of a Supabase backend and a multi-provider AI gateway, the platform maintains a robust architecture capable of supporting its vision of a unified, intelligent workspace.
Sources: README.md:17-25, GEMINI.md:10-20