Skillchain v3.0: 76 Skills, 10 Domains, Separated Architecture
Skillchain v3.0 represents a major milestone with complete coverage across all development domains. This release brings 76 production-ready skills, 12 category orchestrators, and a cleaner separated architecture.
What's New in v3.0
76 Production-Ready Skills
We've grown from 29 skills in v2.1 to 76 complete skills covering every aspect of modern software development:
| Domain | Skills | Examples |
|---|---|---|
| Frontend | 15 | theming-components, visualizing-data, building-ai-chat |
| Backend | 14 | api-patterns, databases-relational, auth-security |
| DevOps | 6 | testing-strategies, building-ci-pipelines, implementing-gitops |
| Infrastructure | 12 | operating-kubernetes, writing-infrastructure-code, configuring-nginx |
| Security | 7 | architecting-security, implementing-compliance, implementing-tls |
| Developer Productivity | 7 | designing-apis, building-clis, debugging-techniques |
| Data Engineering | 6 | architecting-data, streaming-data, optimizing-sql |
| AI/ML | 4 | implementing-mlops, prompt-engineering, evaluating-llms |
| Cloud | 3 | deploying-on-aws, deploying-on-gcp, deploying-on-azure |
| FinOps | 2 | optimizing-costs, resource-tagging |
12 Category Orchestrators
Each domain has its own specialized orchestrator, plus cross-domain handlers:
Single-Domain Orchestrators (10):
- frontend.md - React, Vue, Svelte, component libraries
- backend.md - APIs, databases, authentication
- devops.md - CI/CD, testing, GitOps
- infrastructure.md - Kubernetes, IaC, networking
- security.md - Security architecture, compliance, hardening
- developer.md - API design, CLIs, SDKs
- data.md - Data pipelines, warehouses, streaming
- ai-ml.md - MLOps, prompt engineering, evaluation
- cloud.md - AWS, GCP, Azure patterns
- finops.md - Cost optimization, tagging strategies
Cross-Domain Orchestrators (2):
- fullstack.md - Coordinates frontend + backend skills
- multi-domain.md - Handles complex cross-cutting requests
12 Blueprints for Common Patterns
Pre-configured skill chains for rapid development:
# Infrastructure & Security
/skillchain:start k8s-deployment
/skillchain:start ci-cd-pipeline
/skillchain:start security-hardened-api
# Data & Analytics
/skillchain:start data-pipeline
/skillchain:start ml-pipeline
/skillchain:start dashboard
# Cloud & Operations
/skillchain:start cloud-native-app
/skillchain:start observability-stack
/skillchain:start cost-optimized-infra
Separated Directory Architecture
v3.0 introduces a cleaner separation between commands and data:
~/.claude/
├── commands/
│ └── skillchain/ # Exposed as /skillchain:* commands
│ ├── start.md # /skillchain:start
│ ├── help.md # /skillchain:help
│ ├── blueprints/ # /skillchain:blueprints:*
│ └── categories/ # /skillchain:categories:*
│
└── skillchain-data/ # NOT exposed as commands
├── registries/ # 10 domain-specific YAML registries
└── shared/ # Common resources
Benefits:
- Cleaner command namespace (no
/_registryclutter) - Faster command discovery
- Better organization of data files
- Easier to maintain and extend
New Command Pattern
The main command is now /skillchain:start instead of /skillchain:
# New pattern (v3.0)
/skillchain:start dashboard with charts
/skillchain:start REST API with postgres
/skillchain:start kubernetes deployment
# Help and documentation
/skillchain:help
# Direct blueprint access
/skillchain:blueprints:dashboard
/skillchain:blueprints:ci-cd
# Direct category access
/skillchain:categories:frontend
/skillchain:categories:backend
Domain Highlights
Infrastructure Skills (12)
Complete infrastructure automation coverage:
- operating-kubernetes - Deployments, services, ingress, HPA, operators
- writing-infrastructure-code - Terraform, Pulumi, CloudFormation patterns
- administering-linux - System administration, systemd, package management
- architecting-networks - VPC design, subnets, routing, firewalls
- load-balancing-patterns - Layer 4/7, health checks, algorithms
- planning-disaster-recovery - Backup strategies, RTO/RPO, failover
- configuring-nginx - Reverse proxy, SSL, caching, optimization
- shell-scripting - Bash best practices, error handling, portability
- managing-dns - Route53, zones, records, DNSSEC
- implementing-service-mesh - Istio, Linkerd, mTLS, traffic management
- managing-configuration - Config management, environment variables
- designing-distributed-systems - CAP theorem, consistency patterns
Security Skills (7)
Enterprise-grade security patterns:
- architecting-security - Zero trust, defense in depth
- implementing-compliance - SOC 2, GDPR, HIPAA frameworks
- managing-vulnerabilities - CVE management, patching strategies
- implementing-tls - Certificate management, mTLS
- configuring-firewalls - Network security, segmentation
- siem-logging - Log aggregation, alerting, correlation
- security-hardening - CIS benchmarks, OS hardening
AI/ML Skills (4)
Complete AI development lifecycle:
- implementing-mlops - Training pipelines, model versioning
- prompt-engineering - Chain-of-thought, few-shot, evaluation
- evaluating-llms - Benchmarks, quality metrics, A/B testing
- embedding-optimization - Chunking strategies, retrieval tuning
Installation
The interactive installer now prominently features skillchain installation:
# Clone and run installer
git clone https://github.com/ancoleman/ai-design-components.git
cd ai-design-components
./install.sh
Interactive Menu (v3.0):
What would you like to do?
1) Full Install - Marketplace + all plugins + /skillchain command
2) Install Skillchain - Install /skillchain:start command globally
3) Marketplace + Plugins - Add marketplace + install all plugins
4) Marketplace Only - Just add the marketplace
5) Select Plugins - Choose which plugins to install
6) Uninstall - Remove marketplace and plugins
7) List Plugins - Show available plugins
8) Help - Show all commands
Direct Installation:
# Install everything
./install.sh
# Select option 1
# Install skillchain only
./install.sh commands
Example Workflows
Full-Stack Dashboard
/skillchain:start analytics dashboard with postgres backend
Skillchain automatically:
- Routes to fullstack orchestrator
- Loads creating-dashboards, visualizing-data, databases-relational
- Applies theming-components for consistent styling
- Configures auth-security if needed
Kubernetes Deployment
/skillchain:start deploy my app to kubernetes
Skillchain automatically:
- Routes to infrastructure orchestrator
- Loads operating-kubernetes, writing-infrastructure-code
- Adds observability and security-hardening
- Generates complete k8s manifests
Security-First API
/skillchain:start secure REST API with OAuth
Skillchain automatically:
- Routes across backend + security domains
- Loads api-patterns, auth-security, implementing-tls
- Adds rate limiting, input validation
- Configures proper CORS and headers
Upgrade Notes
From v2.1
If upgrading from v2.1:
- Re-run installer:
./install.sh commands - Update commands: Use
/skillchain:startinstead of/skillchain - Directory changes: Data now in
~/.claude/skillchain-data/ - Preferences preserved: Your saved choices still work
Breaking Changes
- Command changed from
/skillchainto/skillchain:start - Directory structure reorganized (commands vs data separated)
- Registry format updated (domain-specific YAML files)
Performance
- Skill loading: ~2-3 seconds (parallel loading)
- Registry lookup: under 50ms (domain-specific files)
- Path discovery: under 50ms
- Total startup: ~3-4 seconds
What's Next
Future enhancements planned:
- Skill recommendations - AI-suggested skill chains based on project context
- Template library - Community-contributed blueprints
- Skill marketplace - Browse and install community skills
- Project scaffolding - Generate complete starter projects
- Cross-domain intelligence - Smarter skill combinations
Resources
- Skillchain Documentation
- Skills Overview - All 76 skills
- Creating Skills - Contribute new skills
- Full Changelog
- GitHub Repository
76 skills. 10 domains. One command.
/skillchain:start [describe what you're building]

