Hello world, again
Who I am, how I work now, and what this site is for.
I’ve been writing software professionally since the early 2000s. Almost all of it in .NET — Web Forms, WCF, MVC, Web API, Blazor, the whole lineage. Twenty-five years of shipping things for a living.
About two years ago I started working seriously with AI coding assistants, and sometime in the last year the work changed underneath me. Not the tools — the work itself. The way I think about building software is different now, and I don’t think it’s going back.
This site is where I write about what that looks like in practice.
What I actually do
At my day job I’m a Senior Software Engineer for a supply-chain management firm. I run an AI Developers meeting, where we talk about AI Coding Assistants and other developer-specific topics.
Outside of work I’m building side projects, taking courses, and doing the kind of agentic development I don’t always get to do on the clock. This site is one of those projects.
How I work now
My daily workflow is Claude Code and Cursor, usually together. Claude Code for the heavy structural work — multi-file refactors, new feature builds, anything that needs to hold a lot of context at once. Cursor for the edit-level work where I want inline completions and quick iterations.
The interesting part isn’t the tools themselves. It’s what changes about how you think when you have them.
I plan differently. I write more context documents and fewer comments. I spend more time on verification and less on first drafts. I think about problems in terms of what the model needs to know to solve them — context engineering, not just prompt engineering. The skill that transfers isn’t “write better prompts.” It’s knowing what to put in the context window and what to leave out.
The thing I keep coming back to: speed without verification is just faster failure. The models are fast. Making sure they’re right is the actual work now.
What this site is for
I’m building fortworthdev.com as a resource for two kinds of developers:
-
Developers learning to use AI coding assistants well. Not “here’s how to install Copilot” — the kind of practical knowledge that comes from using these tools on real codebases every day. What works, what breaks, what habits to build.
-
Developers getting into agentic and workflow development. MCP servers, autonomous pipelines, multi-agent architectures — the stuff that’s moving from research demos to production. I’m learning this by doing it, and the notes go here.
The site itself is an example. It was designed and built almost entirely through AI coding agents — the terminal layout, the component architecture, the deployment pipeline. The workflow behind it will be the first case study once I write it up.
Why “in public”
I’m not an expert in agentic development. I’m a 25-year veteran of shipping .NET software who is learning this stuff in real time, and I think that perspective is more useful right now than expertise.
Most of the writing in this space comes from people who are either selling a tool, selling a course, or selling themselves as thought leaders. I’m not doing any of those things yet. I just want to document what works and what doesn’t while I’m close enough to the learning curve to remember where the sharp edges are.
If something I write saves you a day of fumbling with context windows or agent loops, the site did its job.
The resources section is curated the same way — nothing goes in that I haven’t actually used, and every entry says why it earned a slot and what it gets wrong. The blog posts in the writing section are filed under assistants/ or agents/ depending on which track they serve.
I’ll write about what I build, what I break, and what I learn. If you’re on the same path, follow along.