Working with AI coding tools

Most guides tell you to use AI tools. This one tells you how. Practical methodology for experienced developers: spec-driven development, agent setup, documentation verification, and learning faster without the noise.

AI dev workflow

What to put in your Cursor rules (and what not to)

A five-tier taxonomy for organizing Cursor rules: constitution, rules, skills, engineering docs, and CI. More rigorous than anything currently documented.

Coming soon

Agentic development workflow: a practical framework for teams

How to structure a team that uses AI coding agents day-to-day - shared knowledge, agent instruction files, human workflow, design control, and governance.

Coming soon

CLAUDE.md best practices: how to set up AI coding agents properly

What goes in CLAUDE.md, AGENTS.md, and GEMINI.md - and how to write agent instructions that actually work across different tools.

Coming soon

Agent-agnostic development setup: one repo, any AI

How to configure a repository so it works well with Claude, Cursor, Gemini, and whatever comes next - without rewriting your setup every six months.

Coming soon
Documentation gap technique

How I use Claude to analyze open source code instead of reading documentation

Why reading documentation is often the wrong starting point, and how to use AI to go straight to what the code actually does.

Coming soon

How to verify what open source documentation actually says

A repeatable technique for checking whether documented features work as described - before you build on top of them.

Coming soon

How I found a bug in Apache Iceberg's upsert implementation before it reached production

The docs claimed upsert worked. An AI scan of the source code said otherwise. A real case study in why documentation gaps matter.

Coming soon

Debezium JDBC Oracle: when the docs don't apply

The official Debezium docs were written for Postgres. Here is how AI helped find the Oracle-specific gap and implement a working fix.

Coming soon
Learning methodology

How to use AI coding tools as an experienced developer (not a beginner)

AI tools are often taught as if everyone is starting from scratch. Here is how experienced developers should actually use them.

Coming soon

Stop watching long YouTube videos. Use NotebookLM instead.

How to use NotebookLM to summarize videos, generate mind maps, and extract what you actually need in a fraction of the time.

Coming soon

The gap analysis method: how to learn only what you need with AI

How to identify exactly what you don't know, fill only that gap, and stop wasting time on things you already understand.

Coming soon

How to learn faster with AI tools

Practical techniques for using AI as a learning accelerator - not just a code generator - when picking up a new technology or skill.

Coming soon

Stay ahead.

One email when something worth reading is published.

No spam. Unsubscribe any time.