I built UseWhisper.dev — an AI code reviewer that analyzes your code diffs, PRs, or snippets and returns review feedback instantly. It runs in the browser with no signup required, and is meant to give developers quick second opinions on logic, style, security, and best practices.
What it does:
Paste a diff, GitHub PR link, or code snippet
Get line-by-line intelligent feedback
Suggestions on readability, errors, anti-patterns
No login, minimal UI, fast responses
Why I’m posting: It’s still early and likely wrong in many ways. I want honest testing, brutal roasting, and real use-case feedback. Break it, trick it, embarrass it — everything helps.
Questions I care about:
Where does it give useful feedback vs hallucinations?
What common review suggestions does it miss?
UX friction when pasting diffs / navigating comments?
Danger points (wrong advice that would be harmful)?
Where should this fit in a real dev workflow?
If you tweet code at it, torture test async/edge cases, or compare with humans — post results! Thanks for the heat