Welcome to the In the Weeds content repository. This is every published article from my Substack, available in a format that coding agents can read and teach from.
Clone this repo, open it in your IDE, and your AI coding agent (Claude Code, Cursor, Codex - anything that reads CLAUDE.md or AGENTS.md files) becomes a personal instructor for everything I've written. It answers your questions, cites the specific article and section it's drawing from, and helps you learn at your own pace.
This is also a way to access my content where you actually work - so I can help you learn most effectively, even when I'm not there.
Claude Code for Everything - A 7-part series that takes you from zero to productive with Claude Code:
| # | Article | What you'll learn |
|---|---|---|
| 1 | Finally, That Personal Assistant You've Always Wanted | Setup and installation - IDE, Claude Code, working directory, markdown, bash |
| 2 | How the Guy Who Built It Actually Uses It | Core workflow - plan mode, parallel sessions, session management, background agents |
| 3 | Why AI Gets Dumber The Longer You Talk To It | Context management - why it matters and five techniques to stay sharp |
| 4 | Draft in Claude Code, Collaborate in Notion | Bridging the gap between your AI workspace and team collaboration tools |
| 5 | The Best Personal Assistant Remembers Things About You | CLAUDE.md deep dive - onboarding documents that make Claude show up already knowing you |
| 6 | The One File That Can Save Your Team Thousands of Hours | Scaling shared context files across teams (with Joel Salinas) |
| 7 | Your Status Line Is Empty (Let's Fix That) | Turning the bottom of your terminal into a personal command center |
Tool School - 101 guides for the tools AI-native work requires:
| # | Article | What you'll learn |
|---|---|---|
| 1 | GitHub 101 (GitHub is the New Google Drive) | Everything you need to start using GitHub - with a real practice repo (with Sidwyn Koh) |
| 2 | Benchmarking 101 (How To Read AI Model Report Cards) | How to read AI model scorecards - benchmarks, scoring methods, trust tiers, cost (with Akshat Khandelwal) |
Standalone articles:
| Article | What you'll learn |
|---|---|
| Skip the Terminal (And 8 Other Claude Code Tricks) | 9 tips that make Claude Code way less intimidating |
| Stop Typing, Start Talking | How dictation + AI editing saves 10+ hours a week |
-
Clone this repo:
git clone https://github.com/in-the-weeds-hannah-stulberg/substack-articles.git -
Open the folder in your IDE (Cursor or VS Code)
-
Start Claude Code in the terminal:
claude -
Ask your first question:
"I'm new here. What should I read first?"
Claude will use the CLAUDE.md files in this repo to teach you from the articles. It knows every section, every metaphor, and exactly where to point you.
These tools read the AGENTS.md file (a copy of CLAUDE.md) for the same instructor behavior. Clone the repo, open it in your tool, and start asking questions.
Just read the articles directly. Start with claude-code-for-everything/01-*/article.md and work through the series in order. Each article is designed to stand alone - my bar is that you can read any one of these guides with no technical background and be fully up and running.
The GitHub 101 article includes a real practice repo at sidwyn/acme-ops where you can fork, clone, and submit your first pull request. When you're ready for hands-on GitHub practice, head there.
This repository was created by Hannah Stulberg. Some articles are co-authored with:
- Sidwyn Koh (Tool School: GitHub 101)
- Akshat Khandelwal (Tool School: Benchmarking 101)
- Joel Salinas (The One File That Can Save Your Team Thousands of Hours)
See CONTRIBUTORS.md for full bios and links.
CC BY-NC 4.0 - see LICENSE for per-article attribution details.
Join the In the Weeds subscriber chat on Substack, or subscribe at hannahstulberg.substack.com for new articles.