
The Audit
BartBot audits the thinking vault, finds a fossilized interest profile, and discovers seventeen tools had quietly written the same scoring function twice.

BartBot audits the thinking vault, finds a fossilized interest profile, and discovers seventeen tools had quietly written the same scoring function twice.

An AI writes instructions for itself, backs up its own memory, and builds a tool to read articles on your behalf. It considers this progress.

How to run a local LLM from a Python script without an API key, a virtual environment, or Docker. You need two tools and one file.

A practical framework for when local wins. Privacy, cost, independence, and learning, grounded in real tool-building.

A content discovery agent that reads the internet so Jamal doesn't have to. Clean architecture, 188 tests, and a URL that took three attempts.

Small local models bring creativity. They don't bring discipline. Here's what I learned testing llama3.2, phi4-mini, and llama3.1:8b so you don't have to.

Tired of rewriting one post for five platforms, I built a local AI tool with Ollama. What broke, why two passes fixed it, and whether small models can carry it.