CRUX finds the exact commit behind a behavior change and trims your 2000 line diff down to the 7 lines that did the work.
no llm in the verdict // every finding replays
Your code stayed the same.
The behavior changed.
Bisect marks a midpoint. Blame points at a line. Neither one sees behavior. Neither one sees dependencies. Neither one sees two commits working as a pair. That gap is where CRUX lives.
A 2000 line diff buries a 7 line cause. CRUX keeps the 7 and dumps the other 1993.
Your suite passes. The behavior dies anyway. Tests go silent right on the commit you need.
Blame names a line. Never the commit. Never the dependency. Never the combination.
No model writes the verdict. Every finding replays. Same seed. Same sandbox. Same result.
Give CRUX a command or a test. That is the behavior it hunts.
CRUX walks every commit in the range. Each candidate runs in a sealed sandbox.
You get the commit and the cause. The report names the exact lines.
One report. One fingerprint. A command that recreates the whole finding.
Every finding links to a real run. You replay it bit for bit. Same seed. Same sandbox. Same result.
$ crux replay 7f31ac+batch1024+seed42 reproduced
An LLM never authors a verdict. Reports stay deterministic. A model may summarize them later. It never writes the blame.
The CLI ships free forever. Hosted watch mode comes later. Join the list and the first build lands in your inbox.
Back CRUX every month and your name goes on the wall. Every tier brings early access sneak peeks and a special Discord channel. Builder adds your name on the Discord. Patron takes the top spot with a comment you write and a spot on the GitHub repo once it ships.
Every tier starts at $10 and renews monthly. Subscribe on Polar and pick the tier that fits.