53+ min ago   (1182+ words) Introduction: The Paradox of the Amnesiac Assistant It's an experience we've all had: you start a new project with a cutting-edge AI like Cursor, Windsurf, or Copilot. The first few hours are magical. The architecture is clean, the commits are…...

1+ hour, 14+ min ago   (599+ words) This is a continuation of my "Claude Code environment" series. Last time I wrote about making Google Calendar the ground truth for my Vault. This time it's about the timeout hell that hit the same vault-auto-ingest job — and how I…...

1+ hour, 22+ min ago   (125+ words) I've run over 100 post-mortems. The worst ones end with 'Alice will be more careful.' The best ones end with 'we fixed the system.' Here's how you get from the first to the second. Ban these phrases from retros: People make…...

58+ min ago   (467+ words) So the thesis of nano-harness is score-per-line-of-code: the smallest readable harness that still puts up a real number on a real benchmark. Think of Karpathys nanochat, but for agent harnesses —small enough to read end-to-end in one sitting, honest enough…...

53+ min ago   (787+ words) A reader named Ryan left the sharpest comment on my last post. The gist: it was jargon-heavy, kept restating "you have to test the AI's output" in new words, and read more like a pitch deck than a case study....

1+ hour, 9+ min ago   (879+ words) We got the ball started… and honestly, it’s easier to keep pushing that ball than to dive head‑first into another direction. Slightly off‑topic for the exact scope of this series, but hey — when in Rome, right? If you're…...

1+ hour, 29+ min ago   (651+ words) You pin your npm dependencies. You have a lockfile. You review the diff when it changes. Now consider the MCP servers your agent depends on. What pins those? tools/list hands back names, descriptions, and JSON schemas, and your agent…...

1+ hour, 9+ min ago   (340+ words) Most agent tutorials stop at tools: give the model a function schema, hope it calls the right one. That works for get_time and hash_string. It falls apart when the model skips a knowledge search and opens a ticket, or when eighty APIs…...

1+ hour, 11+ min ago   (785+ words) This guide is the checklist I wish every team ran before flipping DNS. It’s framework-accurate for the Next.js App Router, and it works whether your new backend is headless WordPress, a headless CMS, or flat files. Every decision in…...

53+ min ago   (386+ words) Most agent projects do not fail because the first model call is difficult. They become difficult when the prototype needs memory, tools, evaluation, a user interface, documentation that coding agents can navigate, and a repeatable review process. That is the…...