“We bought it because the agents were outrunning review”
Used daily for 6–12 months
Adopting Claude Code doubled our merge volume and our review queue quietly became the constraint. CodeRabbit does the first pass, so people arrive at a PR that has already been summarised and triaged. Time to first human comment dropped from about a day to under two hours.
Pros
The summary alone changes how people approach a big diff. One-click fixes for the routine stuff. It learned our conventions within a month.
Cons
The first fortnight was noisy enough that two engineers muted it. Worth pushing through.
Problems it solves
Fixed the review bottleneck that adopting an agent created.
J
James W.Validated Reviewer
RevOps Manager · SaaS · 51–200 employees
4.0
July 11, 2026
“A good first pass, not a reviewer”
Used daily for 6–12 months
It reliably catches the things a tired human misses at 5pm — unhandled errors, a test asserting nothing, a stray credential. It has no idea whether the change was the right thing to build, and a team that treats its comments as mandatory will waste time arguing with it about style.
Pros
Consistent on the mechanical problems. Genuinely useful summaries on large diffs.
Cons
Occasionally confident about something stylistic. Not a substitute for someone who knows the product.
O
Oliver H.
Agency Owner · Marketing agency · 1–10 employees
4.5
June 24, 2026
“Free on our open-source repos, which sold it internally”
Used weekly for 6–12 months
We ran it on our public repositories for two months at no cost before anyone had to approve a purchase. By the time I asked for budget the team had already decided they wanted it on the private ones. Easiest procurement conversation I have had.
Pros
Free tier on public repos is unlimited and complete. Setup was a GitHub app install.
Cons
Twenty-four dollars a seat on top of our existing AI spend takes some justifying to finance.
Problems it solves
Gave a small team consistent review coverage without hiring a senior reviewer.
