What this is
This project collects experiments around local AI tools: small models, practical agent workflows, and the surrounding scripts that make results easier to repeat.
The writing should stay lightweight. A project page can explain what was tried, what worked, what failed, and what should be revisited without exposing private data or turning the site into a build log.
What belongs here
- Short experiment notes.
- Screenshots or diagrams when they help the reader.
- Plain conclusions and links to public code when there is something worth sharing.
The goal is a durable notebook that is readable by people and easy for search engines to parse.