killdate.devGetting fluent with how LLMs actually work — not just using them, but understanding the principles behind their behaviour — makes everything in this series land better. These are the resources worth spending time on.
12-factor-agents — Principles for building reliable LLM applications. If something is going wrong in your agentic system, one of these factors is usually why.
anthropic-cookbook — Code and guides for building with Claude. Copy-paste snippets for tool use, streaming, vision, embeddings. The fastest path from “how does this API work” to working code.
anthropic-courses — Five structured courses: API fundamentals, prompt engineering, tool use, multi-agent systems, building with Claude. Start here if you’re new to the API.
Brex prompt engineering guide — Production-focused. Written by engineers building with LLMs before it was mainstream. More grounding than tutorial.