Looking for feedback on Drift: useful findings, false positives, and missing repo types

[sauremilk]

Hi all,

I’m looking for honest feedback on Drift, a deterministic analyzer for architecture erosion in Python codebases.

If you try it on a real repository, I’d especially like to learn three things:

1. Which findings felt genuinely useful?
2. Which findings felt wrong, noisy, or unclear?
3. Which repository types, frameworks, or architectural patterns should Drift support better next?

I’m not looking for polished reviews. Even one screenshot, one confusing result, or one false positive is useful.

If you do try it, context that helps a lot:

• the kind of repository you ran it on
• whether the output felt actionable
• whether anything looked misleading
• what you expected Drift to detect but it did not

If your feedback turns into something concrete:

• bugs / false positives / false negatives: please open an issue
• feature ideas: feature request is great
• broader design ideas: feel free to reply here

Thanks. I’m especially interested in feedback from people trying Drift on real codebases rather than toy examples.