Jev for App Metadata Release Decisions
Dragos Roua@dragosroua
Jev is used to determine if a metadata-only app release, updating keywords based on Applyra and App Store statistics, is necessary.
very happy to be in the early cohorts of @typesafeai, testing Jev.
This is my first use case - which counts as a tutorial for myself too, because Jev is a bit difficult to understand if you only interacted with LLMs.
Basically, I'm using Jev to decide, with confidence, if a new metadata-only release for my apps, in which I update keywords, based on Applyra and App Store stats, is necessary or not. This kind of work is time consuming, even after I automated it with my own aso-agent, so depending on that decision, I can either save at least 1-2 hours of work, or push a release that's important for my apps' traffic.
In the playground below I am just setting up the criteria and the type of answers I want. The "state" is now just describing the overall goal, but in the next iteration should be dynamic / scriptable: ideally, I should just look into a folder and see if there's anything pulled yet, and how much of it (I already have a Grok @bot for this, it sends me data every morning from these 2 sources).
Right now, I just got 28% confidence in moving foward, because obviously there is no analytics data in the state. The next choice Jev's highlight, with 78% confidence is that I should actually look at the stats.
This can be all scripted in a more stable workflow, so, once I have the data from Grok Bot, I pass this question again to Jev, and after I record Jev's decision, I either surface an alert saying we're good to move forward, or just trigger the metadata-only release flow.
ngl, this thing is *extremely* useful, once you get out, mentally, from the LLMs "chat mode".
I'll share more, once I get some more stuff packed together.