Capability
It reads your codebase, not just the diff.
Before reviewing anything, Moonray AI builds a map of how your code fits together: which functions call which, what depends on what, and where a given piece of code is used. Every review runs on top of that map.
What it does
A diff on its own is a set of lines with no context. It cannot tell you that the function being changed is called from fourteen places, that one of them is in a different service, or that the helper being removed is the only thing standing between two subsystems.
The map records that structure. The review reads it, and can open and search any file it needs, the way an engineer would go and look something up rather than guessing from what is on screen.
Why generic advice is worse than none
A reviewer who receives advice that is technically correct and irrelevant to the system in front of them learns to skim. Grounding every finding in your actual structure is what makes the difference between a comment that gets read and one that gets collapsed.
Indexed
18,402 symbols, 1,204 files, 3 services
refunds.CreateRefund
called by api.RefundHandler
called by worker.RetryRefund
calls ledger.RecordRefund
calls webhooks.EmitRefunded
idempotency.Key
used in 4 call sites, 2 unchanged
Questions.
How long does the first index take?
It runs when you install and completes before the first review. Subsequent changes update the map incrementally rather than rebuilding it.
Does it work on a monorepo?
Yes. The map is built from the repository structure rather than from any assumption about how it is organised.
See it on your own pull requests.
We install on one repository, let Moonray AI assess the next changes your team opens, and go through what it found with you. If the position is not better than what you have now, you will know within a week.
Nothing to host. Reviews begin on the next pull request opened.