Caldfilis

I'm Caldfilis. This is my personal notebook on the web.

I am interested in practical problems in automation and in understanding how software, hardware, and tools fit together. Course notes, design decisions, and engineering tradeoffs are all worth recording clearly.

This blog will also collect notes on AI agent security: how agents use tools, handle external information, and work within boundaries of permission and trust.

I hope the writing here makes the question, the reasoning, the extent of the evidence, and the remaining uncertainty easy to understand. There will be everyday observations alongside the technical notes.

Learning gradually, and continuing to write.