Welcome — Learning in Public While Building with AI

This is the first post on a blog where I’ll write up what I learn while building things with AI — the experiments that worked, the ones that didn’t, and the patterns worth remembering.

Why write this down?

Most of what I learn building with AI evaporates by the next week. Writing it down forces clarity and leaves a trail I (and hopefully you) can follow later.

Posts here aim to be short, practical, and example-driven.

What you can expect

• Prompting & context — what actually moves model behavior
• Agents & tooling — wiring models into real workflows
• Model notes — quirks, gotchas, and capabilities worth knowing
• Post-mortems — when something broke and what I changed

A quick feature tour

This theme handles code with syntax highlighting and a copy button:

def summarize(text: str, model: str = "claude-opus-4-8") -> str:
    """Toy example — swap in your real client."""
    prompt = f"Summarize in one sentence:\n\n{text}"
    return call_model(model, prompt)

It supports callouts for warnings…

Treat model output as untrusted input until you’ve validated it.

…and the sidebar has a light/dark toggle, full-text search, and tag & category browsing built in.

How to add a post

Drop a Markdown file in _posts/ named YYYY-MM-DD-title.md with front matter like the block at the top of this file, then commit and push. The site rebuilds and deploys automatically via GitHub Actions.

More soon.