~ the stack

Tools and links

Everything we use during the workshop, in one place.

Coding agents

Claude Code

Anthropic's agentic coding tool. Terminal-based, works with Claude models. Our primary demo tool.

Open

OpenCode

Open-source AI coding agent. Works with any LLM provider — OpenRouter, Anthropic, OpenAI, or local models.

Open

Model providers

OpenRouter

Unified API for dozens of models — Claude, GPT, Gemini, DeepSeek, Qwen. One key, many providers.

Open

Hugging Face

Open model hub. Browse, download, and run open-weight models. Home of the Open LLM Leaderboard.

Open

Workshop collaboration

HackMD

Collaborative Markdown editor. Shared notes during the workshop — everyone can edit in real time.

Open

tldraw

Collaborative whiteboard. Sketch workflows, architectures, and pipeline diagrams during exercises.

Open

Research tools

refine.ink

AI-powered peer review. Finds errors in logic, mathematics, and consistency before you submit.

Open

dodo

DuckDB extension for .do pipelines. Show SQL, assert, dodoc — the structured IR we demo in Act 3.

Open

Further reading

Other economists teaching AI workflows. Different emphases, complementary material.

AI Guide for Economists

Adrien Matray's guide. Strong on verification mindset, failure-mode taxonomy, CLAUDE.md customization, and async workflows.

Read

AI for Professionals Who Don't Code

Chris Blattman's site. Covers skills, deep research workflows, the council pattern (parallel critics), and executive-assistant automation.

Read

The AI Economist

Aniket Panjwani's training site. Strong on tool choice (Codex vs Claude Code), git as agent memory, plan-before-delegate, permissions and sensitive-data strategies, and AI budget allocation across PIs and RAs.

Read