Blog
Thinking Out Loud
About AI That Works.
Notes on building, deploying, and enabling real AI systems in real life.
-
June 17, 2026
The 3-Layer Skill Architecture of a Working Agentic OS
If your AI agents are making mistakes and burning through API tokens, you are missing a structured skill architecture. Here is how to build it using Identity, Router, and Knowledge layers.
Read → -
June 10, 2026
Custom API Agents vs. Agentic OS: The Architecture That Saves Your Budget
Building custom AI agents from scratch can drain your API budget in minutes. Here is why you should build an Agentic OS instead.
Read → -
May 29, 2026
How to Safely Use GitHub: A Simple Approach for New Users
Most leaked secrets and hijacked accounts come from a handful of habits new GitHub users skip. Here is the short list that actually protects you.
Read → -
May 28, 2026
Give Your Agent Eyes: Web Automation with Claude CLI and Playwright
Playwright isn't just for testing anymore. When you combine it with a CLI agent, it becomes the agent's eyes and its hands on the web.
Read → -
May 27, 2026
Tokens are the New Fuel: Why Context Engineering Matters in Production AI
Dumping everything into the prompt works in sandbox testing, but burns cash in production. Here is how context engineering saves API costs and speeds up execution.
Read → -
May 27, 2026
The End of UI: Why AI Agents Live in the Terminal
We spend millions building beautiful visual dashboards for AI agents. But agents don't want a UI - they live in the CLI.
Read → -
May 23, 2026
Don't Chase AI Brains - Build Your Operating System
The constant flood of new AI models is exhausting. Stop trying to keep up with every update and build an architecture that doesn't care who is winning.
Read → -
May 21, 2026
Stop Treating AI Agents Like a Web Chat
Most people treat AI agents like a web chat. It's costing them performance, speed, and burning their daily token limits.
Read → -
May 17, 2026
The Claude Code Command Nobody Is Running (But Should Be)
There are 60+ built-in commands in Claude Code. /insights is the one that shows you how you work - not just what you built.
Read → -
May 16, 2026
Claude Code Cheat Sheet: The Slash Commands
Every built-in slash command in Claude Code, organized by situation - with recommendations on which ones to use every day.
Read → -
May 11, 2026
Build a Social Carousel Pipeline Without a Design Tool
How I use Claude Code and Playwright to go from topic to published carousel in under 10 minutes - with a copyable setup skill.
Read → -
May 10, 2026
The LLM Wiki: Give Any AI Agent an Infinite Memory
Andrej Karpathy posted this concept and it went viral. Here's what it actually means in practice - and how to build one.
Read → -
May 8, 2026
Give Claude Code Eyes: Playwright for Live UI Observation
Claude Code writes frontend code without seeing the result. Playwright fixes that - and it changes how fast you can build.
Read →