Cheezy World

In this post I will be using Stride and Claude Code to add a Resources section to the Stride UI. Through a series of videos you will see me go through the entire lifecycle of adding this feature to Stride.

Brainstorming with Claude

I start with a concept. I want to add a new section to Stride where users can go to find answers to questions about Stride. I hope to include a lot of images and videos. The video below shows the output of a Brainstorming session I had with Claude. The session lasted about 5 minutes. I would expect a Product Owner and Developer to pair on this activity in a team setting.

In my pursuit of trying to get the most out of my AI Agents I have spent a lot of time exploring how one can deliver requirements in such a way that they can be understood by Humans and provide the details that will make the Agent implementation successful. Knowing that there is typically a pause between when the requirements is established / captured and when the Agent starts working on it, I wanted to make sure that much of the context and details the Agent had during the planning session was captured in a way that it can be use later during implementation.

Does AI work better with some languages than others?

A lot of developers identify with the language they program in. How often do you hear “I’m a Java developer” or “I’m a Python developer”?

I’ve had several conversations with other developers about this topic. Some believe that certain languages are better suited for AI-assisted development than others. Others think that the choice of language has little to no impact on the outcome. I’m of the opinion that language does matter but this is just based on my personal experience that is limited to three languages (Java, TypeScript, and Elixir) in four different contexts.

Sharing what worked

I have often wondered why some companies have had great success adopting AI for software development while others have had the opposite experience. Some people talk about AI “Slop” meaning very bad code written by AI. At the same time many of us see AI producing code that is high quality, well factored, and well tested. Is it possible that the way one approaches this technology has an impact on the outcome?