200+ production-ready AI agent skills for Claude Code and GitHub Copilot. Agentic architecture with 93% token reduction. Now includes comprehensive safety guardrails for damage control.
npm install tech-hub-skills200+ production-ready AI agent skills for Claude Code and GitHub Copilot.
Features a hierarchical multi-agent system with a Brainstorm → Plan → Implement workflow, 95% token efficiency, advanced domain expertise, and comprehensive safety guardrails for damage control.
``bash`
npx tech-hub-skills install
`bash`
npx tech-hub-skills install --copilot
_This creates .github/copilot-instructions.md with all expert skills._
---
95% token reduction with on-demand skill loading and intelligent coordination.
1. Orchestrator Agent (visible) - The master coordinator. Analyzes requests, brainstorms approaches, and creates execution plans.
2. 5 Lead Agents (visible) - Domain experts for AI/ML, Platform, Security, Data, and Product development.
3. 25 Specialist Agents (internal) - Deep technical specialists (e.g., AI Engineer, MLOps, SRE) loaded dynamically by Leads.
- Brainstorm: Understands requirements, constraints, and risks BEFORE acting.
- Plan: Scans registries to select ONLY the needed skills (typically 3-7 per task).
- Implement: Executes step-by-step with validation checkpoints and adaptive planning.
---
- 200+ Skills: LLMs, RAG, MLOps, DevSecOps, Lakehouse, Cloud (AWS/Azure/GCP), and more.
- Lazy-Loading: Significant token savings via internal skill registries.
- Security First: Built-in PII detection, security hardening, and compliance automation.
- Safety Guardrails: File deletion protection, database safety, credential protection, automatic backups, and audit logging.
- Cost Aware: AI/ML cost optimization and efficient tool usage.
Route directly to the master coordinator or domain leads:
`bashRecommended entry point (Full Workflow)
/orchestrator "Build a customer churn prediction model"
$3
Copilot automatically applies expert knowledge via instructions. Reference roles in comments to steer:
python
Using AI Engineer approach for RAG pipeline
def build_rag_system():
passApply Security Architect best practices
def handle_user_upload(file_data):
pass
``| Lead | Domain | Specialists |
| ----------------- | ---------------------- | --------------------------------------------------------- |
| AI/ML Lead | AI, ML, Data Science | AI Engineer, ML Engineer, Data Scientist, MLOps |
| Platform Lead | Infrastructure, DevOps | DevOps, SRE, Platform Eng, Network, Docker, Cloud, FinOps |
| Security Lead | Security, Compliance | Security Architect, Compliance Officer, Security Hardener |
| Data Lead | Data Engineering | Data Engineer, Data Governance, Database Admin |
| Product Lead | Product Development | Product Designer, Frontend/Backend Dev, QA, Tech Writer |
- Full System: AGENTS.md
- Safety Guardrails: SAFETY-GUARDRAILS.md
- GitHub Copilot Guide: GITHUB_COPILOT.md
- Changelog: CHANGELOG.md
MIT