<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Atlas Review Blog</title><description>Engineering notes, benchmarks, and lessons from building Atlas Review.</description><link>https://loomtide.ai/atlasreview/</link><item><title>CodeGraph Cut Atlas Review Cost by 68%</title><link>https://loomtide.ai/atlasreview/blog/codegraph-cut-review-cost/</link><guid isPermaLink="true">https://loomtide.ai/atlasreview/blog/codegraph-cut-review-cost/</guid><description>We added CodeGraph to Atlas Review runs and reran the benchmark across six real OSS pull requests. Mean review cost fell 67.6%, tokens dropped 70.3%, and no repo regressed on cost.</description><pubDate>Tue, 02 Jun 2026 00:00:00 GMT</pubDate></item><item><title>How AI Reviews Code: Why Verification Beats Diff-Only Review</title><link>https://loomtide.ai/atlasreview/blog/how-ai-reviews-code/</link><guid isPermaLink="true">https://loomtide.ai/atlasreview/blog/how-ai-reviews-code/</guid><description>AI code review works best when the model can inspect context, verify suspicions, and turn vague comments into grounded findings.</description><pubDate>Fri, 22 May 2026 00:00:00 GMT</pubDate></item></channel></rss>