Blog
Notes, histories, and deep dives from the Project Sandstar team.
Workflows: The Missing Half of AI-Assisted Building Automation
An assistant that knows Axon syntax perfectly will still build you a broken Spark rule. Not because it cannot write code — because nobody told it the procedure. Twenty-three markdown files are fixing that, and they need you.
Local GPU, OpenRouter, or Both: Routing Every Inference Role Independently
Five roles, four policies, one shared budget — and a compatibility gate that refuses to let a cloud embedder quietly poison an index a local GPU built.
The Sidecar Fleet and the RLM: Gathering Evidence Before Answering
A retrieval model that calls search tools in a loop until it actually has the code, and a fleet agent that tells the server what hardware is awake. Two pieces, one answer.
One Question, Four Languages: Cross-Stack Code Intelligence with Fantom MCP Server
A dropdown in a Vue component, a Fantom weblet, a SOX command over UDP, and a Sedona controller. They are one feature and four languages — and until now, four separate searches.
Teaching an AI a Language It Has Never Seen: Inside Axon MCP Server
Most language models have never read a line of Axon. Without grounding they either refuse to help or invent a function that does not exist. Here is the knowledge layer that closes the gap.
History of Project Sandstar — Part 1: What is Project Sandstar?
The origin story behind an open-source initiative to change how we manage smart buildings — and why we named it after Sedona, Arizona.