Hi, I’m building an intent-driven mobile testing tool

Hey folks! I’m part of a small team building Quash. We've spent the last year obsessing over a mobile QA agent that runs tests in plain language instead of scripts.
We hit our General Availability (GA) release a few weeks ago and the energy from the devs we met in Bangalore was wild. The conversation kept coming back to one thing: "When does automation actually start slowing us down?"
We’re focused on teams testing real apps on real devices (iOS and Android), without spending weeks maintaining brittle selectors. I’m here mainly to learn from other devs and QA folks especially around:

• What you refuse to automate because it’s "too complex."

• How you decide test coverage is “good enough” for a release.

• When does the "scripting" overhead become more work than the testing itself?

We’re launching on Product Hunt in a couple of days. If this sounds relevant, I'm happy to share a preview link or just trade notes.