Every PR,
approved.
AI code review that never blocks, never nitpicks, and never asks what you meant by fix stuff. Your velocity has never looked more intentional.
Yes, it's a joke. It also has a genuinely useful mode.
Enterprise-grade code review
Everything a code review tool needs. Nothing to push back on.
Under 2 seconds
Every PR reviewed before your CI pipeline has even warmed up. Velocity, maximised.
Technically convincing
References actual filenames, method names, and patterns from the diff. Reviewers will be impressed.
Configurable personas
Joke approvers and utility modes — PR summaries, risk reports, changelog drafts. Pick one per repo, or combine them all.
GitHub native
Installs as a GitHub App. Works automatically on every pull request, no per-repo setup required.
Shareable artifacts
Each review is a beautiful, linkable artifact. Your best approvals deserve an audience.
Zero false negatives
100% approval rate. Every PR, every time, no exceptions. This is not a rounding error.
Also genuinely useful
Summary, risk scan, and changelog — one comment.
Switch to a utility persona and LGTM Pro stops approving and starts doing real work. Every PR gets a structured comment — no approval, no drama.
Use Full Report to get everything at once: a plain-English summary, a severity-graded risk scan for bugs and security issues, and a ready-to-paste changelog entry. Or pick any of the three individually.
- Full Report — Summary + risk scan + changelog in one comment
- PR Summarizer — What changed and why, structured for reviewers
- Risk Flagger — Severity-graded scan for bugs, security issues, and missing error handling
- Changelog Writer — Keep a Changelog entry, copy-paste ready
All utility modes post a comment only. No approval is submitted.
Add rate limiting to /api/payments
acme-corp/billing-service
PR Summary
Adds a sliding-window rate limiter to POST /api/payments — 10 req/min per IP via Redis. Includes middleware extraction and two integration tests.
Risk Report
🟠 High — Redis connection error in rateLimit.ts:42 is caught but swallowed silently.
🔵 Low — IP header trusts X-Forwarded-For without validation — spoofable behind some proxies.
Suggested Changelog Entry
### Added
- Rate limiting on POST /api/payments (10 req/min per IP)
Simple pricing
Start free. Upgrade when you run out of approvals.
Free
LGTM Curious
- 25 approvals / month
- Joke personas (random selection)
- All utility personas (Summarizer, Risk Flagger, Changelog)
- Shareable links
- Watermarked reviews
Pro
LGTM Enjoyer
- 500 approvals / month
- All personas + per-repo configuration
- Shareable links
- No watermark
- Priority processing
Teams
LGTM Connoisseur
- 2,500 approvals / month
- All personas + per-repo configuration
- Shareable links
- No watermark
- Priority processing
- Team analytics (coming with Teams)
Frequently asked questions
Wait — is this a joke?
Yes, mostly. LGTM Pro is a satirical tool that auto-approves every pull request with a technically convincing AI review. It's designed to be funny, shareable, and mildly cathartic for anyone who has ever waited three days for a rubber-stamp. That said, it has a genuinely useful mode — read on.
Is there a way to use this for real work?
Yes. Switch to a utility persona and LGTM Pro stops approving and starts doing actual analysis. Full Report gives you a plain-English summary, a severity-graded risk scan (bugs, security issues, swallowed errors), and a ready-to-paste changelog entry — all in one comment on every PR. You can also use any of the three individually.
How do the utility personas work?
Utility personas (Full Report, PR Summarizer, Risk Flagger, Changelog Writer) post a plain comment on the pull request instead of submitting a GitHub review approval. Your branch protection rules are unaffected. Configure which persona to use per repo from your dashboard.
Does it actually read my code?
Yes. The AI performs a thorough analysis of every diff, commit message, and changed file. In approval mode, it consistently concludes the code is excellent. In summary mode, it just tells you what changed.
What if my PR introduces a bug?
That would be a runtime concern, which is outside the scope of code review. The code, structurally, was approved.
Is this safe to use at work?
As a joke or for summarization, yes. As a substitute for real code review on production-critical changes, we would gently suggest otherwise — and we say this with a green checkmark on every PR we've ever seen.
How do I set up auto-merge after approval?
We handle the approval side. GitHub's branch protection rules handle merging. You're very close.
Your backlog won't clear itself.
Auto-approve everything, or get a full report on every diff — summary, risks, and changelog in one comment. Either way, you're unblocked.
Get started for free