alt Hacker NewsNot specific to PartsBox, but we use Inventree (open source similar to PartsBox) and self host it. Over the past few months we noticed certain pain points in our workflow. Rather than looking for a new tool, we used Claude Code to write some backend services and some frontend modifications. Took 2 days of tinkering. Has easily saved that much time since we implemented it.
While rolling the whole solution with an AI agent is not practical, taking a open source starting point and using AI to overcome specific workflow pain points as well as add features allows me to have a lower cost, specifically tailored solution to our needs.
Replies
First, I'll second that I've applied agentic LLMs to an open source project to fix bugs and forcibly coerce it to act in ways that the maintainer may or may not approve of. It has been remarkablely effective, so long as I'm willing to apply patches or maintain a fork of the project (trivial, since this particular open-source project is abandoned anyway).
That said, the act of doing this- using LLMs to dominate somebody's legitimately intelligent and unique work- feels not only discourteous, but worse, like it's a short-term solution.
I'm convinced that it's a short-term solution NOT because I don't think that LLMs can continuously maintain these projects, but because open-source itself is going to be clawed back. The raison d'être of open-source is personal pride, hiring, collaboration, enjoyment, trust, etc. These motivations make less sense in an LLM-fueled world.
My prediction is that useful and well maintained open-source projects like we're hijacking will become fewer and far between.
Won't this breakdown if you need to pull new changes from the original project?
I would include this in the "spreadsheet" metaphor. I do not know your use case, so please don't take this as addressed to you specifically, but I found that there is a learning/complexity problem: many people do not realize there is much more to inventory and production management than it seems. It might seem easy to AI-code something, only to find out later that things could have been done much better.
This is actually a serious problem for me: my SaaS has a lot of very complex functionality under the hood, but it is not easily visible, and importantly it isn't necessarily appreciated when making a buying decision. Lot control is a good example: most people think it is only needed for coding batches of expiring products. In reality, it's an essential feature that pretty much everyone needs, because it lets you treat some inventory of the same part (e.g. a reel) differently from other inventory of this part (e.g. cut tape) and track those separately.
AI-coding will help people get the features they know they need, but it won't guide them to the features they don't know they could use.