CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Ground rules (non-negotiable)

Never delete data or code — do not remove, overwrite, or destructively change any file, directory, or content unless the user gives express permission for that specific deletion.
Stay inside this directory — do not read, write, or operate on anything outside this repository directory unless the user gives express permission.

What this is

Personal academic website for Raymond He (economics, UT Austin), hosted at https://raymondxuanhe.github.io via GitHub Pages. It is a customized fork of the academicpages Jekyll template (itself derived from Minimal Mistakes). Most of the directory structure and template plumbing is generic academicpages; the sections below focus on what is specific to this site or otherwise non-obvious.

Commands

bundle install                              # install Ruby/Jekyll deps (delete Gemfile.lock if it errors)
bundle exec jekyll serve -l -H localhost    # serve + live-reload at localhost:4000
docker compose up                           # alternative: serve via the provided Dockerfile
npm run build:js                            # rebuild assets/js/main.min.js after editing assets/js/**

Note: _config.yml is not reloaded by jekyll serve — restart the server after editing it.

There is no test suite. GitHub Pages builds and deploys automatically on push to master.

Content model

Content lives in front-matter markdown/HTML, not code. Pages are in _pages/; the four Jekyll collections (configured in _config.yml) are _publications/, _talks/, _teaching/, and _portfolio/. Site-wide config and the top navigation are in _config.yml and _data/navigation.yml. Layouts (_layouts/), partials (_includes/), and styles (_sass/) are the generic template internals — change these only for structural/theming work.

Publications and talks can be bulk-generated from TSV via the scripts in markdown_generator/ (publications.py/talks.py read *.tsv and emit collection markdown). This is optional; entries are also written by hand.

Files and downloads

files/ holds all downloadable assets served at /files/..., organized by purpose (CV/, research/, teaching/, benchmarks/, julia_files/, Bio/). The CV is authored in LaTeX at files/CV/raymond_he_CV.tex and committed alongside its compiled .pdf (plus latexmk build artifacts). The job-market paper PDF lives under files/research/jmp/.

scripts/update_cv_json.sh + scripts/cv_markdown_to_json.py convert a markdown CV into _data/cv.json. Heads up: the shell script expects _pages/cv.md, which does not exist in this repo — this pipeline is inherited from the template and is not the live CV workflow (the CV is the LaTeX file above).

Julia / computational benchmarks

Unusually for a Jekyll site, this repo also carries a Julia project (Project.toml, Manifest.toml) used to produce the economics computation benchmarks and charts shown on the site. The Julia sources live in files/julia_files/ (e.g. opt_growth.jl, generate_benchmark_charts.jl); results are surfaced via _pages/computational_benchmarks.*. This is separate from the website build — you do not need Julia to serve or deploy the site.