Blog
Notes on code review & security
Product launches, API design, and practical guides for teams shipping with AI code review.
Benchmarks
AI
Engineering
We ran Patchlight through Martian's open-source Code Review Bench — 50 real PRs, 137 human-verified findings, one judge for every tool. Here are the numbers, the methodology, and where we're strong.
9 min read
Read postProduct
AI
Why we built automated PR/commit reviews and a scheduled security monitor, billed with simple pay-as-you-go dollar pricing.
6 min read
Read postEngineering
AI
How we keep model choice flexible so you can route reviews to Gemini, Claude, or OpenAI per task.
5 min read
Read postSecurity
Privacy
Per-repo control, one-click deletion, and a hard no-training guarantee.
4 min read
Read post