Software
Research Tools, MCP Servers & Open Source
Overview
I develop tools and software to support research in psychometrics and quantitative psychology. Many of my recent projects are developed with the assistance of Claude Code, Anthropic's AI coding assistant.
MCP Servers
Model Context Protocol (MCP) servers for macOS automation, built with Swift:
che-things-mcp
Things 3 automation — 37 tools for complete task management control
che-ical-mcp
Apple Calendar & Reminders management via EventKit
che-apple-mail-mcp
Email workflow automation for Apple Mail
che-claude-desktop-mcp
Claude Desktop control via Accessibility permissions
Research Software
NSQL (Natural Structured Query Language)
A framework for translating natural language to SQL queries
NSQL is a domain-specific language framework that bridges natural language and database queries, making data analysis more accessible to non-programmers.
- Natural language query parsing
- SQL code generation
- Database query optimization
- EBNF grammar specification
Tech Stack: R, YAML, EBNF
Thurstonian IRT Models
R implementations for comparative judgment analysis
Tools for fitting Thurstonian IRT models to analyze preference data from:
- Paired comparisons
- Likert scales
- Joint models combining both
Polychoric Correlation Toolkit
Exploring identifiability conditions
Implementation of methods from my Psychometrika paper on polychoric correlation identifiability with latent elliptical distributions.
Skills & Technologies
Software Development
| Technology | Description |
|---|---|
| AI Agent Integration / Claude Code / MCP | Developed 18 MCP servers for macOS automation, academic workflows, and AI tool ecosystems |
| R / Shiny / Python | Statistical computing, interactive dashboards, data science and automation |
| Swift (iOS/macOS) | Native Apple platform app development, MCP server implementation |
| DuckDB / PostgreSQL / Supabase | Analytical databases, production databases, backend-as-a-service |
| TypeScript / Next.js | Frontend web applications, React-based UIs |
Statistical & Psychometric Methods
- R packages: tidyverse, lavaan, mirt, brms, ggplot2, shiny
- Statistical software: R, SAS, Python, MPLUS, SPSS, JASP
- Specializations: IRT, SEM, HLM, longitudinal data analysis, paired-comparison models
Languages
- Mandarin Chinese: Native
- English: Fluent (reading/writing), Intermediate (speaking)
- Japanese: Reading proficiency