I build AI architecture systems, agentic workflows, and creative intelligence products.

AI architecture · agent runtime · memory systems · multi-model tooling · creator workflows · learning systems

frankx.ai · GenCreator · LinkedIn · YouTube · Ecosystem Map

My work sits in a few connected layers:

This is not a random repo collection. It is an architecture stack: learn the method, apply it with evidence, operate the agents, and build creative products on top.

• AI Architect Academy — Learn by building with a coding agent in real labs.
• AI Architect — A gated architecture lifecycle for systems that call a language model.
• AI Architect Guide 2026 — The working field guide, claim ledger, and authority-boundary lab.
• Starlight Intelligence System — Memory, assessment, coordination, and governance for agentic systems.
• Agentic Creator OS — Skills, commands, and agents for creators and builders.
• Arcanea — Creative intelligence for worldbuilding, learning, publishing, and imagination.

• Fix a broken RAG pipeline — Debug chunking, retrieval, and context assembly.
• Build a multi-agent system — Coordinate specialist agents around a shared goal.
• Build an MCP server — Learn tool integration with a real TypeScript server.
• OCI AI architecture skills — Community coding-agent patterns for enterprise OCI work.

• Skills: Claude Code · Codex · Cline
• Reference architecture: OCI AI Architect · OCI GenAI Guides
• Adoption and deployment: OCI CoE Starter Kit · OCI One-Click Stacks
• Cross-cloud architecture: Multi-Cloud AI Architect

These are practical entry points, not blanket claims of production readiness. Always check license, maintenance history, prerequisites, and evidence in each repo before using it in a live environment.

The ecosystem already includes a strong foundation:

• Learning and education — AI Architect Academy, AI Architect, AI Architect Guide 2026
• Agent operating systems — Starlight Intelligence System, Agentic Creator OS, Starlight Agent Skills
• Creator products — Arcanea, Arcanea-Labs, Kura
• Knowledge and research systems — Library OS, Second Brain OS, Research Intelligence OS
• Oracle and OCI resources — OCI AI Architects, OCI AI Architect, OCI GenAI Guides
• Cross-cloud references — Multi-Cloud AI Architect

My honest assessment:

• The architecture and learning layer is strong and coherent.
• The core repos are active and intentionally built around a shared method.
• The biggest opportunity is not new ideas — it is sharper organization, naming consistency, and clearer entry points.
• Some repositories are highly strategic, while others are exploratory or still evolving; that is fine, but they should be flagged more clearly as “core”, “research”, “community”, or “experimental”.
• The big win is a single canonical experience: a root ecosystem page, a consistent template, and a better “start here” journey for new visitors.

1. Standardize every repo landing page around the same pattern:

   • one-sentence purpose
   • quick start
   • “who this is for”
   • architecture map
   • “related repos” bloc
   • active maintenance status

2. Put one canonical “entry point” in each org:

   • README.md for the org
   • ECOSYSTEM.md for the map
   • a pinned architecture diagram or repo matrix

3. Separate the stack by role:

   • Learn
   • Design
   • Build
   • Operate
   • Research
   • Community

4. Keep certain repos as public reference stacks, not everything as a product.

   • Core product repos should be maintained and documented.
   • Research or concept repos should clearly say they are experiments.

5. Add community-friendly quality patterns:

   • standard docs sections
   • issue templates
   • contribution guidance
   • architecture diagrams
   • validation evidence in README
   • licensing clarity

npx skills add frankxai/claude-skills-library
npx skills add frankxai/creator-skills

• Architect lane: MCP, orchestration, model routing, context, and agent operations
• Creator lane: Video, music, images, brand voice, publishing, and creative production

• Agent infrastructure — memory providers, orchestration, governance, evaluation, and portable skills
• Creator systems — AI-native workflows for writing, music, video, design, publishing, and research
• Open intelligence systems — reusable substrates for research, family history, health, marine life, learning, and knowledge work
• Creative products — Arcanea, worldbuilding tools, publishing systems, and interfaces for human imagination
• AI architecture education — practical patterns for production agents and AI Centers of Excellence

I came through enterprise AI and Oracle Cloud solution design, building AI Centers of Excellence, multi-agent systems, and infrastructure that had to work at production scale. I now apply that systems discipline to creator tools, open-source agent infrastructure, and creative intelligence.

• Oracle-certified AI Architect
• 12,000+ AI songs created
• 200+ repositories across four GitHub organizations
• Building in public at frankx.ai

• frankxai — Core account: architecture method, academy, platform tooling, and creator products
• Arcanea-Labs — Creative intelligence platforms and companion systems
• oci-ai-architects — Community practice for enterprise AI on Oracle Cloud and multi-cloud architecture patterns

_{For the full project map, see ECOSYSTEM.md.}