I build systems where humans define intent, AI executes, and engineering judgment governs the result.

My focus is not simply using AI to write code faster.
I am interested in AI-organized software production: multi-agent workflows, agentic developer tools, AI-native infrastructure, and human-in-the-loop engineering systems.

• Agentic Engineering — multi-agent coding workflows, task decomposition, AI code review, verification loops
• AI-Native Infrastructure — MCP servers, skills, hooks, model routing, workflow automation
• Full-Stack Systems — TypeScript-first architecture with React, Next.js, Node.js, Hono, PostgreSQL, Supabase, and Docker
• Human-AI Collaboration — systems where AI handles execution while humans own judgment, architecture, and responsibility

An intelligent CLI router that selects the best AI backend for each task.

Focus: multi-provider routing, fallback handling, cost-aware AI execution

A Claude Code plugin for coordinating complex development work through parallel subagents.

Focus: task decomposition, parallel execution, result merging, automated review

Contributor

A self-hosted AI personal assistant for messaging platforms such as WhatsApp, Telegram, Discord, and Slack.

Focus: AI agents, messaging interfaces, personal automation

Former Core Team

An open-source Trading OS with a pluggable AI brain.

Focus: market data, AI analysis, trading infrastructure

AI Engineering
Claude · GPT · Gemini · LiteLLM · RAG · MCP · Skills · Hooks · Agent Workflows

Full-Stack
TypeScript · React · Next.js · Node.js · Hono · PostgreSQL · Supabase · Docker

Automation
n8n · Windmill · Custom Workflow Engines · Human-in-the-loop Pipelines

• Multi-agent software development
• AI-native developer tools
• Model routing and cost optimization
• AI decision systems
• TypeScript-first AI applications

• Website: os.ripoai.org
• Email: codermagefox@codermagefox.com
• X: @CoderMageFox
• GitHub: @CoderMageFox

Code is becoming abundant.
Judgment is becoming scarce.
I build tools for organizing intelligence.