I'm two weeks behind on the newsletter (/newsletter/), so I was trying to be responsible by resisting the urge to document the success I've had with my current coding agent setup. My self-restraint has paid off, as Peter Steinberger essentially wrote the exact post I was planning (https://steipete.me/posts/just-talk-to-it) to write. There's lots of good nuggets in here, and it's uncanny how many I agree with: 1. I also use Codex CLI (https://developers.openai.com/codex/cli/) (well, this fork (https://github.com/just-every/code)) on a $200 ChatGPT Pro (https://openai.com/index/introducing-chatgpt-pro/) plan. Claude Code was an epiphany, but their models are overrated for the task, whereas GPT 5's variants are more adherent and diligent across the board. OpenAI's usage limits are virtually infinite by comparison, too 2. I run 3-6 agents in parallel (usually up to 3 per project and up to 2 projects at a time). Unlike Peter, it's rare I let two agents edit the same codebase simultaneously. GPT 5's codex-medium variant is so fast that the time-consuming activities are brainstorming, researching, unearthing technical debt, and planning refactors 3. While git worktrees (https://git-scm.com/docs/git-worktree) are a very cool feature, they dramatically slow down code integration with merge conflicts. Additionally, I've found it's hard to avoid API and port conflicts when running numerous development instances simultaneously. And when an environment stops working, agents will silently start coding based on speculation and conjecture. Fast feedback through observable execution of code is the single most important thing , so the risk isn't worth the (marginal) reward 4. Hooks, custom commands, and fancy hacks like coder's undocumented auto-drive mode (https://github.com/just-every/code/blob/main/code-rs/tui/src/chatwidget/auto_coordinator.rs#L158) are nice, but they're no replacement for thinking really hard about what you want But really, the reason I've had so much success with Codex in comparison to Claude is that if you get off your ass and do the hard thinking necessary to arrive at an extremely crisp and well-informed articulation of what you want, why you want it, and what obstacles it will face, today's agents will generally do a really good job.

A recruiter sent me this screenshot of some kind of GitHub profile scraper. Aside from naming me as a "top 1%" JavaScript developer (which I'm not sure is a compliment or a threat…), I just couldn't get over the "active on weekends" checkmark. Lady, on weekends I charge double. 🤌

For context, last week I wrote a post bringing to light a number of things Andre Arko had said and done (/posts/why-im-not-rushing-to-take-sides-in-the-rubygems-fiasco/) in the past as a way to provide some context. Context that might explain why any of the principal actors involved in the RubyGems maintainer crisis (summarized well up to that point by Emanuel Maiberg (https://www.404media.co/how-ruby-went-off-the-rails/)) would take such otherwise inexplicable actions and then fail to even attempt to explain them. Today, Jean shed some light on Shopify's significant investments in Ruby and Rails open-source (https://byroot.github.io/opensource/ruby/2025/10/09/dear-rubyists.html), and it actually paints a picture of corporate investment in open source done right. (Disclosure: I know and am friends with several people who work at Shopify on these teams, and unless they're all lying to me, they sure seem to prioritize their work based on what Ruby and Rails need, as opposed to what Shopify wants.) Jean went a step further by contrasting Shopify's approach with the perverse incentives at play when individuals or groups receive sponsorships to do open source. He also drew a pretty clear line of those incentives playing out based on how RubyGems and Bundler maintainers reacted to Shopify's feature submissions. Read the post, it's good.

Jared Norman (https://jardo.dev/blog?topic=all) interviewed me after he wrote about the order in which programmers (https://jardo.dev/order-driven-development) choose to write their code and I offered this response (https://justin.searls.co/links/2025-07-29-upside-down-development/). In this episode, we touch on this before launching into a more expansive discussion on why the agile movement fizzled out and what we can reclaim from a developer workflow perspective now that we're experiencing our first major market upheaval since then with the rise of coding agents. Appearing on: Dead Code (https://shows.acast.com/dead-code) Published on: 2025-09-09 Original URL: https://shows.acast.com/dead-code/episodes/fear-driven-everything-with-justin-searls