Claude Code

I use Claude Code — Anthropic’s terminal-based AI coding assistant — as a daily thinking partner for empirical research, teaching prep, and writing. The materials below are aimed at finance academics curious about using it for similar work.

Talks

Slides from recent talks on using Claude Code in academic research. All decks are public.

1. Claude Code for Academics: An AI Agent for Empirical Research (3rd APSA Workshop on Quantitative Methods, ANU, June 2026). Refined version of the brownbag below. What Claude Code is, the thinking-partner philosophy, CLAUDE.md as institutional memory, the Referee 2 pattern, cross-language replication, and sandbox safety.
   [Slides (PDF)]

2. Claude Code for Academics (UTS Finance Brownbag). Earlier department-talk version of the APSA deck above.
   [Slides (PDF)]

3. Claude Code Skills for Academics (UTS Finance Brownbag, Part 3). Building reusable Claude Code skills (split-pdf, paper-editor, code-audit, wrap-up, skill-writing-guide), and end-to-end tooling setup.
   [Slides (PDF)]

Open Source

Two public repositories I maintain:

1. Claude-Code-Presentation — a collection of Claude Code tools, personas, and presentations.

2. Claude-Code-Skills-for-Academics — custom Claude Code skills tailored for empirical research: data profiling, paper review, R-code conventions, and more.

Related Reading

Scott Cunningham’s Substack series on Claude Code for empirical research was a big influence on how I started using these tools.

Gen Li and Siyang Liu maintain a curated collection of Claude Code learning resources for economics and finance researchers, continuously updated.