Has anyone shipped with TestSprite on a real codebase?

Been spending more time writing tests than actual code lately. Coverage requirements keep creeping up, and hand-writing edge cases for async flows is exhausting.

TestSprite caught my eye because it claims to generate tests by analyzing your existing code structure — not just stubbing them out. The part that surprised me: it apparently identifies race conditions and boundary conditions that manual writers tend to skip.

Wrote a breakdown of what it actually does vs. the marketing: https://tools.skila.ai/tools/tes...

Anyone shipped with it on a real production codebase? Curious whether the generated tests actually catch real bugs or mostly just inflate coverage numbers.