Skillchain

Skillchain is a guided, step-by-step workflow system that chains Claude Skills together to build full-stack applications. Instead of manually triggering individual skills, skillchain automatically routes to the right orchestrator, loads dependencies, and ensures correct execution order.

What is Skillchain?

The /skillchain:start command is the recommended entry point for using AI Design Components. It provides a conversational workflow that:

• Analyzes your goal from natural language descriptions
• Detects the domain(s) (frontend, backend, devops, infrastructure, security, developer, data, ai-ml, cloud, finops)
• Matches relevant skills using keyword detection and scoring
• Orders skills correctly with automatic dependency resolution
• Applies user preferences as smart defaults from previous sessions
• Loads skills in parallel when dependencies allow
• Guides configuration through targeted questions
• Generates production code with theming, accessibility, and best practices

Why Use Skillchain?

Traditional Approach	Skillchain Approach
User must know skill names	Describe what you want to build
User triggers skills individually	Skills chain automatically
Easy to miss required skills	Correct skill order enforced
No theming consistency	Theming always applied first
Manual component assembly	Assembly skill runs at end
Repeat configuration every time	Preferences saved between sessions

Example Comparison

Traditional (requires knowledge):

"Use the theming-components skill, then designing-layouts,
then creating-dashboards, then visualizing-data..."

Skillchain (just describe your goal):

/skillchain:start sales dashboard with revenue charts

Version 3.0 Features

10 Domains, 76 Skills

v3.0 expands from 2 domains (frontend/backend) to 10 comprehensive domains:

Domain	Skills	Focus
Frontend	15	UI components, forms, charts, dashboards
Backend	14	APIs, databases, messaging, deployment
DevOps	6	CI/CD, Docker, GitOps, incidents
Infrastructure	12	Kubernetes, Terraform, networking
Security	7	Compliance, TLS, firewalls, SIEM
Developer	7	APIs, CLIs, SDKs, documentation
Data	6	ETL, streaming, SQL optimization
AI/ML	4	MLOps, prompts, LLM evaluation
Cloud	3	AWS, GCP, Azure deployments
FinOps	2	Cost optimization, tagging

12 Blueprints

Pre-configured skill chains for common patterns:

Blueprint	Category	Description
dashboard	Frontend	Analytics dashboard with charts & KPIs
crud-api	Backend	REST API with database & auth
api-first	Developer	API-first design with OpenAPI
rag-pipeline	AI/ML	RAG with vector search & embeddings
ml-pipeline	AI/ML	MLOps pipeline with training & serving
ci-cd	DevOps	CI/CD with testing & deployment
k8s	Infrastructure	Kubernetes deployment with Helm
cloud	Cloud	Multi-cloud deployment patterns
observability	DevOps	Monitoring, logging, tracing stack
security	Security	Security architecture & compliance
cost	FinOps	Cost optimization strategy
data-pipeline	Data	ETL/ELT data processing pipeline

When detected, blueprints offer a faster path with optimized defaults (3-4 questions instead of 12+).

User Preferences System

Skillchain remembers your choices between sessions:

# Saved to ~/.claude/skillchain-prefs.yaml
global:
  theme:
    color_scheme: "blue-gray"
    theme_modes: ["light", "dark"]
  frameworks:
    frontend: "react"
    backend: "fastapi"

How it works:

• First run: Answer questions, offered to save preferences
• Future runs: Your saved preferences become smart defaults
• Override anytime: Just provide a different answer

Preference Priority:

1. User's explicit choice (current workflow) - Highest
2. Saved preferences (from ~/.claude/skillchain-prefs.yaml) - Medium
3. Default values (from skill definitions) - Lowest

Skill Versioning

All 76 skills are versioned for compatibility tracking:

# In registries/*.yaml
visualizing-data:
  version: "1.0.0"
  # ...

This enables:

• Compatibility matrix tracking which skills work together
• Changelog tracking for each skill's evolution
• Version pinning for reproducible builds

Parallel Skill Loading

Independent skills are loaded concurrently for faster workflows:

# Instead of sequential:
Step 1: theming-components
Step 2: designing-layouts
Step 3: visualizing-data   # waits for layouts
Step 4: building-tables    # waits for viz

# Parallel loading:
Step 1: theming-components
Step 2: designing-layouts
Step 3 (PARALLEL): visualizing-data + building-tables

Skills are grouped by parallel_group in the registry. Dependencies are automatically resolved to ensure safe parallelization.

Quick Start Examples

# Start Claude Code in your project
claude

# Frontend examples
/skillchain:start sales dashboard with revenue charts
/skillchain:start login form with validation

# Backend examples
/skillchain:start REST API with PostgreSQL
/skillchain:start RAG pipeline with vector search

# DevOps examples
/skillchain:start CI/CD pipeline with GitHub Actions
/skillchain:start GitOps deployment with ArgoCD

# Infrastructure examples
/skillchain:start Kubernetes cluster with Helm
/skillchain:start Terraform infrastructure for AWS

# Security examples
/skillchain:start SOC2 compliance implementation
/skillchain:start security hardening checklist

# AI/ML examples
/skillchain:start MLOps pipeline with MLflow
/skillchain:start LLM evaluation framework

How It Works

1. Parse Goal - Skillchain analyzes keywords in your description
2. Detect Domain(s) - Routes to one of 10 domain orchestrators (or fullstack/multi-domain)
3. Match Blueprint - Checks for pre-configured patterns (12 available)
4. Order Skills - Ensures correct execution order with dependency resolution
5. Load Preferences - Applies saved preferences from previous sessions
6. Guide Configuration - Asks questions for each skill (or uses defaults)
7. Parallel Loading - Invokes independent skills concurrently
8. Generate Code - Produces themed, accessible components
9. Save Preferences - Optionally saves choices for next time

Available Skills (76 Total)

Frontend Skills (15)

Group	Skills
Foundation	theming-components
Data Display	visualizing-data, building-tables, creating-dashboards
User Input	building-forms, implementing-search-filter
Interaction	building-ai-chat, implementing-drag-drop, providing-feedback
Structure	implementing-navigation, designing-layouts, displaying-timelines
Content	managing-media, guiding-users
Assembly	assembling-components

Backend Skills (14)

Group	Skills
Data Ingestion	ingesting-data
Databases	using-relational-databases, using-vector-databases, using-timeseries-databases, using-document-databases, using-graph-databases
APIs & Messaging	implementing-api-patterns, using-message-queues, implementing-realtime-sync
Platform	securing-authentication, implementing-observability, deploying-applications
AI/ML	ai-data-engineering, model-serving

DevOps Skills (6)

writing-dockerfiles, testing-strategies, building-ci-pipelines, implementing-gitops, managing-incidents, platform-engineering

Infrastructure Skills (12)

operating-kubernetes, writing-infrastructure-code, managing-configuration, architecting-networks, load-balancing-patterns, managing-dns, implementing-service-mesh, administering-linux, configuring-nginx, shell-scripting, planning-disaster-recovery, designing-distributed-systems

Security Skills (7)

architecting-security, implementing-compliance, managing-vulnerabilities, implementing-tls, configuring-firewalls, siem-logging, security-hardening

Developer Productivity Skills (7)

designing-apis, building-clis, designing-sdks, generating-documentation, debugging-techniques, managing-git-workflows, writing-github-actions

Data Engineering Skills (6)

architecting-data, streaming-data, transforming-data, optimizing-sql, secret-management, performance-engineering

AI/ML Skills (4)

implementing-mlops, prompt-engineering, evaluating-llms, embedding-optimization

Cloud Provider Skills (3)

deploying-on-aws, deploying-on-gcp, deploying-on-azure

FinOps Skills (2)

optimizing-costs, resource-tagging

Workflow Commands

During a skillchain session, you can use:

Command	Action
back	Return to previous skill
skip	Use defaults for current question
status	Show current progress
done	Finish early with current selections
restart	Start over from beginning
help	Show workflow commands

Next Steps

• Install Skillchain globally or per-project
• Learn usage patterns with examples
• Explore blueprints for fast-track presets
• Understand architecture and how it works internally