Skip to main content
AI Design Components

Updates, releases, and insights about AI Design Components

Skillchain v2.1: Modular Architecture

· 4 min read
Anton Coleman
Anton Coleman
Project Maintainer

Skillchain v2.1 introduces a completely modular architecture with intelligent features that make building full-stack applications faster and more intuitive.

<!-- truncate -->

What is Skillchain?​

Skillchain is an intelligent orchestrator that helps you build full-stack applications by automatically loading the right Claude Skills for your project. Instead of manually selecting skills, just describe what you're building and Skillchain assembles the perfect skill chain.

New Features in v2.1​

1. User Preferences System​

Skillchain now remembers your choices and provides smart defaults:

# Saved to ~/.claude/skillchain-prefs.yaml
frontend_framework: next.js
styling_preference: tailwind
backend_language: typescript
database_preference: postgresql

Benefits:

  • No repetitive questions for similar projects
  • Faster project setup
  • Consistent technology choices across projects

2. Skill Versioning​

All 29 skills now have proper versioning (v1.0.0) with:

  • Changelog tracking - See what's changed between versions
  • Compatibility info - Know what works together
  • Deprecation notices - Plan for future changes

3. Parallel Skill Loading​

Independent skills now load concurrently using parallel_group fields:

skills:
  - name: databases-postgresql
    parallel_group: data
  - name: using-vector-databases
    parallel_group: data
  # Both load simultaneously!

Benefits:

  • Faster skill chain initialization
  • Reduced waiting time
  • Smarter dependency resolution

4. Blueprint System​

Pre-configured skill chains for common patterns:

Dashboard Blueprint:

/skillchain dashboard

Automatically loads:

  • creating-dashboards
  • visualizing-data
  • building-tables
  • theming-components

CRUD API Blueprint:

/skillchain crud-api

Automatically loads:

  • api-patterns
  • databases-relational
  • auth-security
  • observability

RAG Pipeline Blueprint:

/skillchain rag-pipeline

Automatically loads:

  • databases-vector
  • ai-data-engineering
  • model-serving
  • api-patterns

5. Category Orchestrators​

Specialized handlers for different domains:

  • Frontend Orchestrator - React, Vue, Svelte projects
  • Backend Orchestrator - API, database, auth patterns
  • Fullstack Orchestrator - Combines frontend + backend
  • AI/ML Orchestrator - RAG, embeddings, model serving

6. Dynamic Path Discovery​

Works globally from any project directory:

# Works from anywhere
cd ~/my-project
/skillchain dashboard

# Automatically finds skillchain installation
# No manual path configuration needed

How It Works​

8-Step Workflow​

  1. Step 0: Path Discovery - Finds skillchain directory
  2. Step 0.5: Load Preferences - Reads saved choices
  3. Step 1: Goal Input - What are you building?
  4. Step 2: Keyword Detection - Identifies relevant skills
  5. Step 3: Configuration - Asks targeted questions
  6. Step 4: Skill Assembly - Loads skills in parallel
  7. Step 5: Execution - Runs the workflow
  8. Step 7: Save Preferences - Remembers your choices

Example Usage​

Simple Dashboard​

/skillchain dashboard analytics

Skillchain:

  • Detects keywords: "dashboard", "analytics"
  • Loads: creating-dashboards + visualizing-data
  • Asks: Chart types? Data source?
  • Assembles components with your branding

Full-Stack RAG App​

/skillchain I need a RAG chatbot with user auth

Skillchain:

  • Detects: "RAG", "chatbot", "auth"
  • Loads: building-ai-chat + databases-vector + auth-security + api-patterns
  • Asks: Vector DB? Auth provider? API framework?
  • Builds complete authenticated RAG pipeline

Quick CRUD API​

/skillchain rest api with postgresql

Skillchain:

  • Detects: "api", "postgresql"
  • Loads: api-patterns + databases-relational
  • Uses preferences: TypeScript (from last time)
  • Generates API with proper patterns

Architecture Improvements​

Modular File Structure (19 files)​

.claude/commands/skillchain/
├── skillchain.md          # Main router
├── _registry.yaml         # 29 skills with metadata
├── _help.md              # Documentation
├── _shared/              # 9 shared resources
│   ├── theming-rules.md
│   ├── execution-flow.md
│   ├── preferences.md
│   ├── parallel-loading.md
│   └── ...
├── categories/           # 4 orchestrators
│   ├── frontend.md
│   ├── backend.md
│   ├── fullstack.md
│   └── ai-ml.md
└── blueprints/          # 3 templates
    ├── dashboard.md
    ├── crud-api.md
    └── rag-pipeline.md

Benefits of Modularity​

  • Easier maintenance - Update one file, not the whole system
  • Better organization - Clear separation of concerns
  • Extensibility - Add new categories/blueprints easily
  • Token efficiency - Load only what's needed

Performance​

  • Skill loading time: ~2-3 seconds (parallel loading)
  • Preference loading: <100ms
  • Path discovery: <50ms
  • Total startup: ~3-4 seconds (vs 8-10 seconds in v2.0)

Installation​

# From repository root - interactive installer
./install.sh

# Install skillchain commands only
./install.sh commands

Upgrade Notes​

If you're upgrading from v2.0:

  1. Re-run installer to get modular structure
  2. Your preferences will be preserved
  3. All existing projects continue to work
  4. New features available immediately

What's Next​

Future enhancements planned:

  • Template library - Community-contributed blueprints
  • Skill recommendations - AI-suggested skill chains
  • Project scaffolding - Generate starter code
  • Skill marketplace - Browse and install community skills

Learn More​