English · 繁體中文
I build governed LLM-agent systems for flood-risk and water research. My work connects agent reasoning, verification, and physically grounded environments.
Declared Git evidence · local-first generation · Unknown stays Unknown.
| FLOODABM | Agent-based model coupled with catastrophe flood modeling — household flood adaptation in the Passaic River Basin (NJ, 2011–2023) |
LLM Agents · AI Harness / Governed Agents · Agent-Based Modeling · Multi-Agent Systems · Catastrophe Modeling · Decision-Making Under Risk · Human-Environment Interaction
- Leveraging Large Language Models for Agent-Based Simulation of Human-Water System Interactions — Water Resources Research, 62(6), e2025WR042111 (2026) · DOI
10.1029/2025WR042111
| ai-research-skills | 5 plugins · 14 research skills · install in one command |
| agent-collab-skills | Multi-agent orchestration — task splitter, output reconciler, debate, shared memory, acceptance gate |
| zotero-skills | Programmatic Zotero CRUD — search, add, classify, annotate, organize literature |
| research-hub | AI-operable research workspace for Zotero, Obsidian, and NotebookLM — CLI, MCP, REST, dashboard |
| academic-writing-skills | Rigorous academic paper writing, revision, and submission — field-agnostic with per-paper journal overrides |
| codex-delegate | Single-agent delegation skill used by the marketplaces above |
| ai-profile | Local-first AI collaboration analytics — explicit Git provenance in, privacy-safe cards and an interactive provider dashboard out |
Selected merged pull requests:
| ZhuLinsen/daily_stock_analysis | Added Taiwan (台股) market support — market detection & routing, an institutional-flow (三大法人) data fetcher, and TWD currency / dividend fixes · merged PRs |
| langchain-ai/openwiki | Restricted ~/.openwiki permissions on Windows, where chmod is a no-op · #367 |
| BuilderIO/agent-native | Routed AGENT_ENGINE=codex-cli to the Codex runner instead of crashing · #1362 |
| Nanako0129/coralline | Fixed auto-layout over-wrapping — measure display columns, not bytes · #10 |
| awesome-agentic-ai-zh |  7-stage agentic-AI roadmap · trilingual (繁中 canonical / 简中 / EN) · 240+ curated resources, from LLM basics to multi-agent production |
Currently available for advisory work — prompt review / audit and AI agent workflow consulting. As a full-time PhD student my availability is limited, but teams or companies working in these areas are welcome to reach out: 📧 wenyuchiou12@gmail.com
perceive → reason → verify → act · repeat