Welcome to my public website showcasing my experiments and explorations in using Large Language Models (LLMs) for generating code and providing insights! As a solo developer, I've found that LLMs can be incredibly useful tools in my workflow, and I'm excited to share my experiences and findings with you.

• Interested in learning more about how I use LLMs to generate code? Check out the archives of chats on this site, where you can explore examples of conversations and see how LLMs can help streamline development processes.
• Interested in my agentic coding workflows? I keep track of custom general-purpose skills over at tdpeuter/agent-instructions.

In January 2026, I started experimenting with different AI-coding harnesses, such as Google Antigravity and GitHub Copilot (CLI). I especially love to use plan mode before dispatching a fleet of agents (e.g., see my personal Copilot workflow instructions).

This does mean however, that a lot of coding and AI-use is now less explicit, i.e. not "just" a chat in a browser anymore. That makes it harder to share these experiences in the dump-a-chat format, meaning these archives are therefore at risk of becoming stale. Always use your own judgement, please.

If you have any insights you would like to share, don't hesitate to contact me.
- Tibo De Peuter