Repowise PR bot
AI code review that knows your whole repo.
Repowise is a GitHub App that reviews every pull request with AI and, from a persistent index of your repository, checks what diff-only reviewers can't: which callers a change breaks, which changed lines no test runs, security findings on added lines, code health before and after, and dead code. Free checks on public repositories; AI review on Pro or Teams.
Free checks on public repos, no account needed. Private repos need Pro or Teams, free to try.
One PR, end to end
about 2 minutes from push to review
- review-nudge.tsxnew file<Linkhref={href(change.kind === "working_tree" ? "wip" : change.label)}className="font-medium text-[var(--color-accent-primary)] hover:underline">
- Change risk
- riskier than 95% of commits
- Patch coverage
- 68 of 80 lines
- Health of changed files
- 8.3 → 8.2
P1Review link opens the wrong change
Trigger: the checked-out branch is named
wip, or its name is a 40-character sha.review-nudge.tsx:48
- Prompt for your agentCopy
Build the link from changeLinkParam(change.id) instead of change.label.
True positive, fixed in d19394dfa. Added review-nudge.test.tsx covering both names.
- Repowise reviewSafe to merge
The code this pointed at changed in d19394d.
Saved for your team
Saved as a candidate decision. Confirm it on Repowise.
One install
Six checks on every pull request, from one GitHub App.
- AI bug review
- Bugs found on the diff line, each with a prompt for your agent. On Pro or Teams.
- Breaking callers
- Every caller outside the PR that a changed function reaches.
- Coverage on changed lines
- The added lines no test runs, once CI uploads a coverage report.
- Security on added lines
- Secrets and risky calls, flagged on the line that adds them.
- Code health before and after
- The touched files scored on both sides, and the dead code a change leaves.
- Change risk and reviewers
- Risk against your repo's history, and who owns the code it touches.
On every pull request
Blast radius
Every caller a change reaches, from the index. Reviewers suggested from who owns that code.
Code health on every PR
#5127The touched files, scored before and after.
Coverage on changed lines
The lines no test runs, marked, from your CI's coverage report.
Security on changed lines
Secrets and risky calls, flagged on the line.
Findings your agent can fix
#5127Each finding carries a prompt for your agent.
Learns from your replies
A reply on a finding is saved for your team. On Pro and Teams, AI review checks the change against decisions recorded for its files.
Quiet when nothing needs you
One comment, kept current. With AI review, a clean PR gets a short 'safe to merge'.
A merge gate when you want one
The Repowise review check, advisory or required.
Cost per PR
About 15 cents a reviewed PR
Proof
Caught on real pull requests.
- P1 · found, fixed, saved#5127Review link opens the wrong changeFixed in one push with a test, and saved for your team as a candidate decision.
- Pushback#5164Rechecks when you push backThe author replied with the reason. The bot re-read the code, agreed and resolved the thread.
- Full review#5127Every part of a review on one PRThe check, one comment, health 8.3 → 8.2, patch coverage 85% and a reply on every thread.
Pricing
Free PR checks on public repos. AI review from your plan's credits.
- a reviewed PR
- About 15 centsa reviewed PR
- the most one PR can cost
- 60 centsthe most one PR can cost
- per repo per month by default
- $10per repo per month by default
Free
$0
Public repositories
- Free PR checks on every public repository, no PR cap
- Breaking callers, code health and dead code on every PR
- Coverage on changed lines once CI uploads a report
- The merge gate and a public report for every PR
- No model, no credits spent
Pro
AI review$15 a month
or $12 a month billed yearly
- Everything in Free, on private repositories too
- AI review, about 15 cents a reviewed PR
- $5 of AI credits a month included, about 33 reviewed PRs
- Never more than 60 cents a PR, $10 per repo per month by default
Teams
$20 a seat a month
From 3 seats
- Everything in Pro, across the team's repositories
- One organization install for every member
- A shared credit pool, $5 a seat a month
- A $100 a day AI spend cap for the team
FREQUENTLY ASKED
Questions teams ask first
Is the PR bot free for open source?
The checks are. Install the GitHub App on a public repository and every pull request gets breaking callers, code health, change risk and the merge gate, with no account and no PR cap. AI review runs on Pro or Teams.
Does it replace my coverage bot or security scanner?
On pull requests, one install covers both. Upload a coverage report from CI and it marks the changed lines no test runs; security findings show on the lines a pull request adds. Code health and the callers a change breaks arrive in the same comment.
Can my coding agent set it up?
Yes. Point your agent at https://docs.repowise.dev/hosted/bot-setup and it walks the setup and tells you the one link to click: a person still installs the GitHub App on GitHub.
What does AI review cost?
About 15 cents a reviewed PR, paid from credits, so 100 reviewed pull requests come to about $15. Every push that changes code gets the full review while the budget has room. Never more than 60 cents a PR, and each repository has a monthly AI spend limit, $10 by default (about 66 reviewed PRs), that you can raise. Pro is $15 a month with $5 of credits included, about 33 reviewed PRs, and past that credits are pay as you go with no per-seat review fee. When credit runs out, the bot falls back to a light review and says so on the pull request. Add credits for the full review.
Does it comment on every pull request?
It keeps one comment per pull request and edits it on every push, so the thread never fills up. With only the free checks, a pull request with nothing to act on gets no comment at all. AI review, on Pro or Teams, reviews every pull request that changes code: a clean one gets one short comment saying it is safe to merge and what was checked, and the Repowise review check on each push says what is still open.
Does it run a model on Free?
No. On Free the bot runs the risk, coverage and health checks: the comment comes from the index of your repository, built from parsed code, the call graph, git history and a 23-marker health scorer. A model runs only for AI review, on Pro or Teams, where it is on by default for every repository you claim and can be switched back to the free checks per repository.
Does it change my code?
No. Each finding carries a prompt for your coding agent, which makes the fix in your normal workflow. The one change the bot can open itself is a draft pull request that adds the coverage upload to your CI, for you to review.
How does it show coverage on changed lines?
Upload a coverage report from CI and every pull request shows which changed lines a test runs, with an optional Repowise / coverage check that fails below your minimum. The bot can open the draft pull request that adds the upload workflow for you.
What can it access?
It reads contents and metadata, and writes its pull request comment (GitHub stores it as an issue comment) and its check runs. It cannot merge. Contents and workflow write are used only for the one-click coverage setup. A repository file cannot switch AI review on or raise its cap, so a fork cannot turn paid review on.
Do I need a Repowise account?
Not on public repositories. Install the GitHub App and the bot comments on their pull requests. Sign in with GitHub to choose a review mode, turn on the merge gate, or add private repositories, which need Pro or Teams.
Which languages does it understand?
The same languages the Repowise index parses, including Python, TypeScript and JavaScript, Go, Java, Kotlin, C#, Rust, C and C++, Ruby, PHP and Swift: 42 in all, 28 of them parsed to a full syntax tree. The co-change and ownership signals come from git history and work for any file.
How is this different from a diff-only AI reviewer?
A diff-only reviewer sees the lines that changed. Repowise also knows the rest of the repository: who calls a changed function, which files history says belong together, which tests reach the change and how healthy the touched files are. That evidence comes from the index without a model, and AI review builds on it instead of guessing.
Which hosts does it run on?
The bot is a GitHub App for public, private and organization repositories. The Repowise CI gates run on GitLab through the Repowise template, and in any other CI as plain commands.
Can I turn it off for one pull request or one path?
Put [skip repowise] in the pull request title to skip it. Ignore paths such as vendor/ or generated files per repository in your settings or .repowise/bot.yaml.
Install it on the repository you are working on today.
Breaking callers, change risk and code health are free on public repositories, no account needed. Your next pull request shows you what it does.
Already installed? Manage installsReviewing for a whole organization? Talk to us