Further Reading#

📌 TL;DR#

• Start with The Illustrated Transformer and Simon Willison's blog for LLM fundamentals.
• Read Anthropic's engineering blog for practical agent-building insights.
• Study the OWASP LLM Top 10 for security essentials.
• Explore projects like LangGraph or CrewAI to see alternative agent frameworks.
• The field moves fast; focus on building and iterating.

🧠 Understanding LLMs (For Beginners)#

• The Illustrated Transformer by Jay Alammar. The best visual introduction to the core architecture behind modern LLMs.
• A Hacker's Guide to Language Models by Jeremy Howard (YouTube). A practical, code-first approach to understanding and using LLMs.
• Simon Willison's Blog. Excellent, accessible writing on prompt engineering, evaluations, embeddings, and agentic workflows.
• Intro to Large Language Models by Andrej Karpathy (YouTube). A comprehensive overview of the LLM landscape, training, and capabilities.

📄 Papers Worth Knowing#

• Attention Is All You Need (Vaswani et al., 2017). The original Transformer paper that started it all.
• Language Models are Few-Shot Learners (Brown et al., 2020). The GPT-3 paper, demonstrating the power of scale and in-context learning.
• LLaMA: Open and Efficient Foundation Language Models (Touvron et al., 2023). Details on a family of influential open-weight models.

🛠️ Building With Agents#

• Anthropic's Engineering Blog. Contains published prompts, Claude Code documentation, and posts on building effective AI systems.
• OpenAI Cookbook. While focused on OpenAI's API, it contains many transferable patterns for function calling, chat, and agentic loops.
• "Building effective agents" (Anthropic blog). Practical advice on designing reliable agentic workflows.
• Embrace The Red Blog. Focuses on AI security and prompt injection, crucial for understanding agent risks.

🔒 Security for AI Agents#

• OWASP LLM Top 10. The standard list of the most critical security risks for LLM applications.
• Lakera's Gandalf. An interactive game (CTF) that teaches you how prompt injection attacks work by trying to "trick" an AI into revealing a password.
• Prompt Injection Archive by Simon Willison. A curated collection of resources, examples, and discussions about prompt injection.

🧰 Open-Source Projects Worth Exploring#

• Model Providers & APIs: Anthropic, OpenAI, OpenRouter. Compare offerings.
• Agent Frameworks: LangGraph, CrewAI, AutoGen. Explore alternative architectures to OpenClaw's design.
• Self-Hosted Inference (for future local setups): Ollama, vLLM. Tools to run models on your own hardware.

🍓 Hardware & Home Infrastructure#

• Raspberry Pi Official Documentation. The definitive source for your Pi's capabilities and setup.
• Raspberry Pi Cookbook by Simon Monk. A practical guide to Pi projects and Linux system management.
• Tailscale and Cloudflare Tunnel. Tools mentioned in the guide for establishing secure remote access to your Pi without complex port forwarding.

🚀 One Closing Thought#

The AI agent landscape evolves rapidly. Don't try to read everything before you start. Use this guide as a launchpad: build your team, break it, learn why it broke, and iterate. The deepest understanding comes from hands-on practice, not passive consumption.