<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>OneStone AI Blog</title><description>Build-in-public engineering notes from OneStone — a local-first AI agent desktop for Linux and Windows.</description><link>https://onestoneai.com/</link><item><title>LongCat-2.0 Beats GPT-5.5 on SWE-bench Pro — What It Means</title><link>https://onestoneai.com/blog/longcat-2-0-swebench-pro/</link><guid isPermaLink="true">https://onestoneai.com/blog/longcat-2-0-swebench-pro/</guid><description>A Chinese open-source model just outscored OpenAI&apos;s latest on coding. Here&apos;s why it matters for the competition.</description><pubDate>Tue, 07 Jul 2026 00:00:00 GMT</pubDate></item><item><title>Anthropic suspends Fable 5 and Mythos 5 over US export controls</title><link>https://onestoneai.com/blog/anthropic-fable-mythos-export-suspension/</link><guid isPermaLink="true">https://onestoneai.com/blog/anthropic-fable-mythos-export-suspension/</guid><description>How a government compliance order cascades through the entire AI stack—and what it means for builders outside the US.</description><pubDate>Mon, 06 Jul 2026 00:00:00 GMT</pubDate></item><item><title>The Line Between Free and Restricted AI Is Now a Benchmark Score</title><link>https://onestoneai.com/blog/ai-regulation-is-a-benchmark-score/</link><guid isPermaLink="true">https://onestoneai.com/blog/ai-regulation-is-a-benchmark-score/</guid><description>The White House is drafting voluntary release standards for frontier AI. The trigger is not a law — it is a cybersecurity eval score between 70.7% and 96.7%.</description><pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate></item></channel></rss>