The project is called Intent Bus. I built it because I wanted to trigger scripts on my devices from a cloud server without opening ports or setting up Redis for something that runs maybe a few times a day. It is aimed at indie developers and home lab people. The kind of workload it is actually built for is a background script that fires a notification when something finishes, or a Pi that picks up a task when your laptop tells it to. Not high frequency, not mission critical, just reliable enough to trust. What I was curious about was whether SQLite would fall apart under concurrent workers. The assumption is always that it will. With WAL mode and Waitress as the WSGI server it ended up handling 40 concurrent workers at 34 jobs per second with 99% success and no lock contention at all. For something running a few hundred jobs a day that is genuinely more than it will ever need. The actual bottleneck was not SQLite. It was the WSGI layer. Gunicorn on a single thread collapsed under concurrent polling.…