Words & Bytes

Tokenmaxxing is a practice that engineers and large tech companies have stepped away from for a few months now.

I'm of the opinion that parallel and background agents are key to how agentic coding improves our productivity.

I'm excited to announce the launch of a Private Beta (i.e. waitlist) for Branchmore, a tool to help engineers measure and improve how they use AI coding tools.

A few months ago as I worked to reduce my AI coding agents' hallucinations, I had a growing realization that I needed to improve my context management.

Most engineers I pair with can spot an AI coding agent hallucination when it happens.

In Part 1 of this series, I enumerated a few obstacles for engineers taking vibe coding from side projects to production.

Let's say you're an engineering manager, and you're participating in an organization-wide campaign to increase the adoption of AI.

I'm going to share some experiences from a few months ago, about how I expanded the scope of my use of agents from vibe coded apps to working on real world problems in production.

In Part 1 of this series, I noticed something strange while using Claude Code's Max 20x plan: it was the same $200/month as Cursor Ultra, doing the same work, but my Claude Code utilization was stuck at 16% while I had been burning through Cursor's token…

At the end of my last post, I was: stuck at 16% Claude Code utilization on the Max 20x plan, and had just figured out that parallel agents could help me break past that limit, explore git worktrees, and make better use of the $200 plan I had paid for.

While tinkering over the holidays, I remember thinking: "This is so strange!

The technique that helped me jump over the threshold from "coding with AI" to actually "vibe coding" was the use of a plain markdown file.