<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>devtake.dev — #vibe-coding</title><description>Articles tagged vibe-coding on devtake.dev.</description><link>https://devtake.dev/</link><language>en-us</language><item><title>Microsoft tested 19 LLMs as document editors. Even the best ones corrupted 25% of the content.</title><link>https://devtake.dev/article/llms-corrupt-documents-delegation-errors/</link><guid isPermaLink="true">https://devtake.dev/article/llms-corrupt-documents-delegation-errors/</guid><description>The DELEGATE-52 benchmark tests AI editing across 52 professional domains. Frontier models corrupt a quarter of document content over long workflows.</description><pubDate>Sun, 10 May 2026 09:00:00 GMT</pubDate><category>ai</category><category>llm</category><category>ai-models</category><category>benchmarks</category><category>microsoft</category><category>delegation</category><category>vibe-coding</category><author>dieter-morelli</author></item></channel></rss>