VDK Docs

VDK Ecosystem Overview

Complete guide to the VDK three-tier architecture and component integration

VDK Ecosystem Overview

The VDK (Vibe Development Kit) ecosystem is a sophisticated three-tier architecture designed to transform generic AI coding assistants into project-aware development experts through intelligent analysis, curated knowledge, and seamless platform integration.

Three-Tier Architecture

┌──────────────────────┐    ┌─────────────────────────────┐    ┌─────────────────────┐
│   VDK CLI v2.0.1     │◄──►│  VDK-Blueprints Repository  │◄──►│   VDK Hub vdk.tools │
│ (Local Analysis)     │    │ (Knowledge Base)           │    │ (Web Platform)      │
│                      │    │                            │    │                     │
│ • Project Scanning   │    │ • 109 Expert Blueprints    │    │ • Blueprint Catalog │
│ • Technology Detection│    │ • Schema v2.1.0           │    │ • 7-Step Generator  │
│ • Pattern Recognition │    │ • 7 Categories             │    │ • Team Collections  │
│ • Blueprint Fetching  │    │ • GitHub Webhooks          │    │ • Real-time Sync    │
│ • Platform Adapters   │    │ • Community Contributions  │    │ • Analytics         │
└──────────────────────┘    └─────────────────────────────┘    └─────────────────────┘
         │                                 │                                 │
         ▼                                 ▼                                 ▼
┌──────────────────────┐    ┌─────────────────────────────┐    ┌─────────────────────┐
│   AI Assistant       │    │   Universal Blueprint       │    │   Team Collaboration│
│   Platforms          │    │   System                    │    │   & Sharing         │
│                      │    │                            │    │                     │
│ • Claude Code        │    │ • Platform-Specific        │    │ • Public Collections│
│ • Cursor AI          │    │   Adaptation               │    │ • Usage Analytics   │
│ • Windsurf           │    │ • Schema Validation        │    │ • Best Practices    │
│ • GitHub Copilot     │    │ • Content Optimization     │    │ • Community Curation│
└──────────────────────┘    └─────────────────────────────┘    └─────────────────────┘

Component Deep Dive

Tier 1: VDK CLI (Local Engine)

Version: 2.0.1
Requirements: Node.js ≥22.0.0
Package: @vibe-dev-kit/cli

The VDK CLI serves as the intelligent local analysis engine that understands your project's unique characteristics and generates appropriate AI assistant configurations.

Core Capabilities

Project Analysis Pipeline:

  1. ProjectScanner: Traverses codebase with .gitignore respect and file categorization
  2. TechnologyAnalyzer: Detects 20+ technologies, frameworks, and build tools
  3. PatternDetector: Advanced naming convention and architectural pattern recognition
  4. RuleGenerator: Synthesizes analysis results using remote blueprints

Blueprint Integration:

  • Remote-First Architecture: Fetches blueprints from GitHub (entro314-labs/VDK-Blueprints)
  • Dynamic Discovery: Runtime discovery of available categories and commands
  • Light Templating: $\{variable} substitution for project-specific customization
  • Relevance Scoring: Sophisticated algorithms match blueprints to project characteristics

Platform Deployment:

  • Universal Compatibility: Simultaneous configuration of multiple AI platforms
  • Specialized Adapters: Platform-specific format generation
  • Confidence Scoring: Integration reliability assessment (low/medium/high)
  • File Structure Generation: Follows each platform's specific requirements

Tier 2: VDK-Blueprints Repository (Knowledge Base)

Location: GitHub (entro314-labs/VDK-Blueprints)
Schema: v2.1.0
Total Blueprints: 109

The centralized repository serves as the authoritative source for all AI assistant blueprints, providing expert-curated content with version control and community collaboration.

Blueprint Categories

CategoryCountDescriptionExamples
Core4Fundamental AI behavior patternsAgent behavior, code quality, security
Languages6Programming language guidelinesTypeScript, Python, Swift, C++20
Technologies26Framework and tool patternsReact 19, Next.js 15, Supabase, Tailwind
Stacks6Multi-technology combinationsNext.js + Supabase, React Native
Tasks54Executable development workflowsCode review, security audit, refactoring
Assistants7AI platform configurationsClaude, Cursor, Windsurf, Copilot
Tools3Development tool integrationsFile operations, command execution

Blueprint Format

---
title: "Blueprint Title"
description: "Detailed description of the blueprint's purpose"
category: "technology/framework"
complexity: "simple|medium|complex"
compatibility: ["claude-code", "cursor", "windsurf", "copilot"]
file_patterns: ["*.ts", "*.tsx", "*.js"]
dependencies: ["react", "@types/react"]
version: "2.1.0"
auto_apply: true
priority: 100
tags: ["frontend", "react", "typescript"]
---

# Blueprint Content

Structured instructions for AI assistants, including:
- Context about the technology or pattern
- Specific coding guidelines
- Common patterns and best practices
- Error prevention strategies
- Platform-specific optimizations

Tier 3: VDK Hub (Web Platform)

URL: https://vdk.tools
Stack: Next.js 15, React 19, Supabase, TypeScript 5.8
Features: Real-time sync, team collaboration, analytics

VDK Hub provides a comprehensive web interface for blueprint discovery, custom package generation, and team collaboration.

Core Features

Blueprint Catalog:

  • Browse 109 blueprints with advanced filtering
  • Search by technology, complexity, compatibility
  • Preview blueprint content and metadata
  • Download individual blueprint packages

7-Step Generator Wizard:

  1. Project information (name, description)
  2. AI assistant platform selection
  3. Technology stack configuration
  4. Programming language preferences
  5. Development tools selection
  6. Environment details
  7. Preview and package generation

Team Collaboration:

  • Personal blueprint collections with sharing capabilities
  • Public collection URLs for team distribution
  • Usage analytics and adoption tracking
  • Real-time synchronization via Supabase subscriptions

Integration Architecture

AI Platform Compatibility Matrix

The VDK ecosystem supports comprehensive AI assistant platform integration through specialized adapters:

PlatformDirectoryFormatFeaturesConstraints
Claude Code.claude/Memory files + commandsMemory hierarchy, slash commandsOptimized chunking
Cursor AI.cursor/rules/.mdc (YAML + Markdown)File pattern matching, auto-completionGlob patterns
Windsurf.windsurf/rules/XML memoriesWorkspace awareness, memory optimization6K character limit
GitHub Copilot.github/copilot/JSON guidelinesRepository settings, review integrationRepo/org level

Data Flow Patterns

CLI → Repository Integration:

  • GitHub API fetching with optional authentication
  • Incremental updates and local caching
  • Rate limiting and error recovery
  • Blueprint currency tracking

Hub → Repository Integration:

  • GitHub webhooks for real-time updates
  • Database synchronization via Supabase
  • Content validation and schema compliance
  • Version control and change tracking

Cross-Component Syncing:

  • CLI and Hub maintain consistent blueprint versions
  • Real-time notifications of available updates
  • Automatic refresh of AI platform configurations
  • Conflict resolution and error handling

Performance Characteristics

CLI Performance

  • Project Analysis: Sub-second scanning for typical codebases (<10k files)
  • Blueprint Fetching: ~2-3 seconds for full blueprint refresh
  • Configuration Generation: ~500ms for multi-platform deployment
  • Memory Usage: ~50MB during analysis, ~10MB at rest

Hub Performance

  • Search Response: <200ms for blueprint search with full-text indexing
  • Package Generation: 2-5 seconds for custom package creation
  • Real-time Updates: Instant synchronization via Supabase subscriptions
  • Global Distribution: CDN-optimized content delivery via Vercel

Repository Performance

  • Content Delivery: GitHub CDN with global distribution
  • Webhook Response: <5 seconds for update propagation
  • API Rate Limits: 5000 requests/hour (authenticated), 60/hour (anonymous)
  • File Size Optimization: Blueprints average 2-8KB each

Security & Compliance

Data Protection

  • User Isolation: Row Level Security (RLS) in Supabase database
  • Authentication: OAuth flows with GitHub integration
  • API Security: Rate limiting and secret key protection
  • Input Validation: Schema enforcement and content sanitization

Content Security

  • Repository Access: GitHub token-based authentication with scope restrictions
  • Webhook Verification: Signature validation for repository updates
  • Content Validation: Blueprint schema compliance checking
  • Version Control: Full change tracking and rollback capabilities

Privacy Considerations

  • Local Analysis: Project scanning happens locally (no code uploaded)
  • Anonymous Usage: CLI can function without authentication
  • Data Retention: Hub analytics anonymized after 90 days
  • Third-party Access: No blueprint content shared with external services

Future Roadmap

CLI Evolution (Q2 2025)

  • VDK Hub API integration for cloud features
  • Enhanced project analysis with dependency graph mapping
  • Team configuration sharing and synchronization
  • Performance optimization for large codebases (>100k files)

Hub Evolution (Q2-Q3 2025)

  • Advanced blueprint recommendation engine
  • Enterprise features with organization management
  • Mobile-optimized progressive web app
  • Integration marketplace for third-party blueprints

Repository Evolution (Q3-Q4 2025)

  • Multi-language blueprint support
  • Automated quality scoring and validation
  • Community voting and rating system
  • Blueprint versioning and compatibility tracking

Getting Started

Quick Setup Workflow

# 1. Install VDK CLI
npm install -g @vibe-dev-kit/cli

# 2. Initialize in your project  
vdk init --interactive

# 3. Verify setup
vdk status --check-integrations

# 4. Validate configurations
vdk validate

# 5. Keep updated
vdk update --verbose

Alternative: Web-Based Setup

  1. Visit https://vdk.tools
  2. Complete the 7-step generator wizard
  3. Download platform-specific packages
  4. Follow integration guides for your AI assistant

Community & Support


The VDK ecosystem represents a comprehensive solution for making AI coding assistants project-aware through intelligent analysis, curated knowledge, and seamless platform integration. The three-tier architecture ensures scalability, reliability, and extensibility while maintaining focus on developer productivity and practical utility.