name: "Zhongyue Lin (Leo) | 林中岳"
location: "China"
working_on:
  - "Codex / Claude Code skill ecosystems"
  - "Multi-model agent teams and AI workflow infrastructure"
  - "Knowledge Hub: Obsidian-based course notes, research notes, and WeChat archives"
  - "Applied ML and data-analysis portfolio projects"
research_bridge:
  - "History and institutional design -> multi-agent orchestration"
  - "Behavioral economics -> agent routing and evaluation"
  - "Knowledge management -> durable AI memory workflows"
languages: ["中文", "English"]

I build practical AI tooling around a simple question:

How do we turn messy human workflows into reliable, inspectable, reusable agent systems?

我关注的是把真实工作流沉淀成可复用的智能体系统：能调度模型、能整理知识、能留下上下文，也能在需要时被人工审查。

2026-06-10

• 🛠️ Upstream CodexBar contributions — three fixes merged into steipete/CodexBar: a Claude web session-recovery action (#1378), a Doubao false-100%-usage fix (#1383), and a macOS 26 menu-tracking fix that defers copy-icon work off the NSMenu tracking loop to kill the beachball (#1389).
• 🧩 OpenCLI adapter work — multiple adapters merged into jackwener/OpenCLI: a manus.im read-only adapter (#1837), Gemini status / history / detail / read commands (#1838), plus new kimi, qoder, trae-solo, antigravity, and codex adapters with conversation-management commands.
• 📐 Math-foundation notes — chapter-level bilingual notes landing in Knowledge-Hub: a full Elements of Information Theory walkthrough (17 chapters + index) and the start of a CS-DIY/数学基础 track (MIT 18.06 linear algebra, convex optimization, deep learning).
• 🏛️ csdiy-deep-enhance mother skill — froze the CMU 15-445 deep-enhancement methodology (3-phase rebuild: link repair → per-lecture enrichment → auxiliary notes) into a reusable mother skill, so any course can be upgraded to the same standard with one trigger.

2026-06-03

• 🏛️ skills-master — promoted the local skill index to a callable Skill Strategist mother skill: 12 disclosure-style categories, 7 enforced workflows (create / merge / archive / audit / integrate-external / placement-decision), 3 templates, and a pre-flight router. Skills can now be authored, merged, archived, and integrated under a single governed flow.
• 🔥 grimoire-skill v2 (in progress) — proposed refactor from a PDF/video tool into a general-purpose Skill Forge Engine with input adapters for webpage, Obsidian vault, GitHub repo, and audio, auto-classified into the 12 skills-master categories. Tracked in PR #16 (open); the shipped repo is still the MinerU document-parsing skill.
• 📚 Product-methodology stack (6 books → 18 skills) — Perri (strategy), Cagan (quarter), Torres (week), Fitzpatrick (sentence), Olsen (process), Nika (AI-PM). All distilled via the book-distill-pipeline mother skill into a uniform 3-skill-per-book structure plus Obsidian notes (~14,900 lines under Books/).
• 🧹 Skill-graph consolidation — pruned 632 local skills to 601 via 13 deletions, 5 cluster merges, and 2 new meta indexes (full archive kept for rollback).

The skill ecosystem is being consolidated around one public canonical repository:

productivity        Obsidian, NotebookLM, Google Workspace, Knowledge Hub
development         MCP, frontend, repo design, D3, R analytics, SQL workflows
documents           PDF, DOCX, PPTX, XLSX, MinerU, document coauthoring
research            scientific skills, STEM modeling, history notes, macro research
ai-collaboration    multi-provider orchestration, agent teams, skill discovery
product-management  Lenny-inspired PM frameworks and playbooks
design              themes, CLI demos, visual assets
system              proxy, CDP, SVN, internal comms, OSS contribution

Governance is handled by two mother skills:

• skills-master — the strategist: classifies, merges, archives, and audits skills under one governed flow.
• grimoire-skill — the engine: turns source material (currently PDF/DOC/PPT via MinerU) into draft skills; a v2 multi-adapter Skill Forge Engine is proposed in PR #16.

Standalone skills kept separate because they still have independent value:

I work on patterns for routing tasks across multiple model families and agent roles:

• role-based agent teams
• model/provider aliases
• smart task routing by skill, language, and intent
• inter-agent messaging and review loops
• durable context handoff between humans and models

Related repositories:

My favorite ideas come from treating old institutions as tested coordination systems.

Blog entry point: leolin990405.github.io/blog

Main index: udacity-masters-ai-projects

Selected projects:

Older course repositories remain useful as project records, while the portfolio index is the preferred entry point.

I am gradually reducing repo sprawl:

Principles:

• one canonical home per workflow
• no secrets or auth material in public repos
• private repos for operational scripts with local credentials
• public repos for reusable frameworks, skills, and portfolio artifacts
• README files should tell users where the maintained surface now lives

• GitHub: @LeoLin990405
• Blog: leolin990405.github.io/blog
• LinkedIn: Zhongyue Lin

If one of my projects is useful, a star is always appreciated.