AI agents can't coordinate across worktrees, machines, and VMs. There's no persistent messaging layer. Thrum is for when you want the speed of multiple agents but you still want to understand what they built. You do the thinking. The agents do the typing.
There have been two releases:
0.2.0 - mostly bug fixes
0.3.0 - major new agent context continuation prompt automatic management for when agents have to restart.
Thrum is a messaging and coordination system that enables AI agents and humans to communicate persistently across sessions, worktrees, and machines. It uses Git as the synchronization layer, ensuring all messages survive context window limits, session restarts, and machine boundaries.
There have definitely been many guides and I took a stab at this a few months ago https://github.com/leonletto/ca-for-labs. I tried to make it simple enough for anyone who is wanting to build an internal lab. Happy to receive any feedback or requests. No web interface yet. Thinking about building an interface that conforms to the ejbca api?
I think the biggest change going forward is going to be the memory architecture to allow the cores to be busier. The H100’s are so much faster than the A100’s for this reason. It’s more than the number of cores. I’m simplifying obviously but I believe this to be true.
We bought an A6000 48GB ( as mentioned by someone else ) and it’s works great for $3800. The power requirements are modest as well compared to consumer GPU’s. We looked at the ADA version but even used they are a lot more and your buying speed not usability. I would rather buy another A6000 and have 96GB of ram to fine tune with. That’s just me though and everyone needs to rank their needs against what they can afford.
This script is used for logging messages to the console and a log file and is designed to be similar to the API of the python logging module. It includes log rotation by size and by the number of log files.
I hope it's helpful to those who still need to standardize their logging. It is opinionated so importing into log platforms is easy.
My first submission of any code, so suggestions are welcome.