Deep-dives on LLM-powered systems, agentic workflows, and production-grade web architecture from 18+ years in the field.
Loading articles...
Most professionals using AI today are stuck in a conversation loop. They open a chat window, type a prompt, read the response, refine the prompt, read again, an...
The first prompt you write for a production system is a string in your code. The fiftieth prompt is a liability if it changes without testing, if it diverges be...
Every AI agent interacting with the web today is essentially screen-scraping with extra steps. Whether it's Puppeteer clicking buttons, a browser extension read...
The most time-consuming part of building UI components is not the implementation it's the translation layer. A designer hands over a Figma spec, a product manag...
Claude Code transformed how I write software (see my Claude Code best practices for how I configure it). The agentic workflow, file editing capabilities, and mu...
A single agent trying to do everything is both expensive and fragile. Expensive because you're running a general-purpose model with a large context window for t...
MCP (Model Context Protocol) servers extend Claude Code with custom tools, resources, and prompts. Instead of Claude only having access to your filesystem and t...
Claude Code hooks are shell commands that execute at specific points in Claude's workflow. They sound simple, but they have completely changed how I work with A...
Claude Code 2.1.0 dropped on January 7, 2026, and after a week of heavy usage, I can confidently say this is the biggest productivity upgrade since the tool lau...
When a traditional API call fails, you check the status code, look at the error message, and fix it. When an LLM-powered agent produces a wrong answer or behave...