A curated collection of resources for AI-assisted development. These are the docs, tools, and ideas I keep coming back to while building with Claude Code.
Getting Started with Claude Code
- Claude Code Documentation — The official reference. Start here for installation, configuration, and core concepts.
- Claude Model Card — Model capabilities, context windows, and which model to use when
- Prompt Engineering Guide — How to write prompts that get better results. Applies to Claude Code commands, skills, and agent definitions.
Building Agents & Workflows
- Building Effective Agents — Anthropic’s research on agentic AI patterns. The foundation for skills, chains, and multi-agent workflows.
- Anthropic Cookbook — Practical examples and guides. Useful for understanding tool use, structured output, and API patterns.
- Claude Code on GitHub — The CLI tool’s source code. Read the issues and discussions for real-world usage patterns.
The Stack
The tools I use daily, and why:
- Supabase — PostgreSQL with Row Level Security, Auth, and Edge Functions. The database layer for the entire Magic Platform.
- Vercel — Three-tier deployment pipeline (feature branches → staging → production). Preview deployments for every push.
- Linear
— Project management via MCP integration. Every
/startand/committalks to Linear automatically. - Hugo — Static site generator powering this blog. Fast builds, markdown-native, good theme ecosystem.
- PaperMod — The Hugo theme this site runs on. Clean, fast, good search support.
Open Source
Projects from the blog that you can explore:
- Adventures in Claude — This site’s source code. Hugo + PaperMod + Vercel auto-deploy.
- Freshell — Terminal multiplexer built with Claude Code. Fork of a Rust project, contributed upstream.
Perspectives
Writing that shaped how I think about AI-assisted development:
- The End of Programming as We Know It — Tim O’Reilly on AI’s impact on software development
- AI-Native Development — The Pragmatic Engineer on AI tools for developers
- llms.txt Standard — Making websites AI-readable. A small spec with big implications for how AI agents interact with the web.
This page is updated periodically. Have a suggestion? Open an issue .