There wasn't much skill involved, but if there's credit to be had for giving Claude a few screenshots, API docs and iterating on the design while doing human product testing, then I guess I'll take it!

I haven't shared it because it's rough (you know what vibe code looks like underneath!) and I only implemented the features I needed. So there's no external IFTTT-like hooks, it probably doesn't internationalize well, and I didn't need collaboration features. I also don't want to be on the hook for supporting a codebase that isn't the core of what I do. My real goal was to see if I could implement a replacement for myself in less than the cost of one annual seat, even at API pricing.

But I'm surprised to see Harvest is still in the news about this... so, maybe! Thanks for the encouragement!

No worries, and you're welcome! My ask for you to share wasn't to have to provide support (as an open source project maintainer, I completely understand), but I am gathering examples of this type of build mode, I think there is great value in building a positive feedback loop where it becomes easier to build replacements when the rug is pulled. This potentially "outruns the enshittification" if you will, a meta version of open source that sits on top of the building via tokens.

> My real goal was to see if I could implement a replacement for myself in less than the cost of one annual seat, even at API pricing.

Indeed. Yes we can, and we can continually improve upon this process. Onward! Thanks again for sharing!