name: "Zhongyue Lin (Leo) | 林中岳"
location: "China"
role: "Multi-AI Systems Architect & Data Analyst"
education:
degree: "Data Science & Analytics"
interests:
- "Multi-AI Orchestration & Agent Teams"
- "LLM Agent Frameworks"
- "Data-Driven Decision Making"
current_focus:
building: "Agent Teams for Claude Code Bridge"
creating: "xiaotie - Lightweight AI Agent Framework"
maintaining: "10+ Claude Code Open Source Skills"
researching: "Multi-Agent Simulation (CivAgent)"
analyzing: "Data Pipelines with Python / R / SQL"
languages: ["中文 (Native)", "English (Fluent)"]
fun_fact: "I route 9 AI providers through a single Gateway" |
|
Multi-AI Architect — I build systems that make 9+ AI providers (Claude, GPT, Gemini, DeepSeek, Kimi, Qwen...) collaborate as unified agent teams. From smart task routing to inter-agent communication pipelines, I design the infrastructure for AI-to-AI collaboration.
Data Analyst — I transform raw data into actionable insights using Python, R, and SQL. Building end-to-end data pipelines from collection to visualization.
Open Source Creator — I build and maintain the largest Claude Code Skills ecosystem with 10+ packages, and contribute to multi-AI collaboration infrastructure used by developers worldwide.
|
Multi-AI Collaboration Infrastructure 多 AI 协作基础设施
My Contributions:
|
Lightweight AI Agent Framework 轻量级 AI Agent 框架
Key Features:
|
||||||||||||||
|
Multi-AI Desktop Platform 智能多 AI 协作桌面平台
Key Features:
|
The Largest Claude Skills Ecosystem 最大的 Claude Code 技能生态
|
||||||||||||||
|
57-Civilization AI Multi-Agent Simulation 57 文明 AI 多智能体模拟
|
AI Usage Stats for macOS macOS AI 使用统计工具
|
┌──────────────────────────────────────────┐
│ User / Claude Code CLI │
└──────────────────┬───────────────────────┘
│
┌─────────────▼─────────────┐
│ Smart Router (--auto) │
│ keyword + skill matching │
└─────────────┬─────────────┘
│
┌──────────────────▼──────────────────────┐
│ Claude Code Bridge (Gateway) │
│ aliases | teams | routing | messaging │
└──────────────────┬──────────────────────┘
│
┌────────┬────────┬───────────┼───────────┬────────┬────────┐
▼ ▼ ▼ ▼ ▼ ▼ ▼
┌────────┐┌────────┐┌────────┐┌─────────┐┌──────┐┌────────┐┌────────┐
│ Claude ││ GPT ││ Gemini ││ Kimi ││ Qwen ││DeepSeek││ iFlow │
│ (c) ││ (a) ││ (b) ││ (d) ││ (h) ││ ││ │
└────────┘└────────┘└────────┘└─────────┘└──────┘└────────┘└────────┘
Communication Patterns:
--to agent : Directed message (A → B)
--broadcast : Fan-out to all team members
--chain a|b|c : Sequential pipeline with context passing
--auto : Smart routing by content analysis
人类 5000 年的政治制度史是一部组织架构设计模式库——57 种政体，6 种 AI 编排模式。 「历史不只是过去的事，它是组织智慧的活化石。」
- 问题的提出：为什么 AI 编排需要历史学？
- 六种编排模式的类型学
- 唐代三省六部——质量门控的经典实现
- 明代双轨制——N-Version Programming
- 雅典民主——分布式知识管理系统
- 波斯总督制——最终一致性
- 跨学科桥梁、技术实现与反思
Keywords: Multi-Agent Systems, Political Science, Institutional Design, LLM Orchestration
