<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Agentic Madness</title><description>Working with AI agents: experience, ideas and experiments.</description><link>https://agentic-madness.com/en/</link><language>en</language><atom:link href="https://agentic-madness.com/en/rss.xml" rel="self" type="application/rss+xml"/><item><title>What Benchmarks Really Measure</title><link>https://agentic-madness.com/en/posts/what-benchmarks-really-measure/</link><guid isPermaLink="true">https://agentic-madness.com/en/posts/what-benchmarks-really-measure/</guid><description>How seven LLM benchmarks work: tasks, verification, metrics and run conditions. What release-table percentages mean, and how to compare models on your own work.</description><pubDate>Mon, 02 Nov 2026 00:00:00 GMT</pubDate><category>benchmarks</category></item><item><title>What Is Jev and What Is It For?</title><link>https://agentic-madness.com/en/posts/what-is-jev-and-what-is-it-for/</link><guid isPermaLink="true">https://agentic-madness.com/en/posts/what-is-jev-and-what-is-it-for/</guid><description>How Jev turns free-form text into scores for preset options, where it fits in software, and what comparisons with Rev, Unsloth and Microsoft-Decision-1 show.</description><pubDate>Sun, 18 Oct 2026 00:00:00 GMT</pubDate><category>models</category></item><item><title>Two Ways to Make an Agent Read Logs</title><link>https://agentic-madness.com/en/posts/two-ways-to-make-an-agent-read-logs/</link><guid isPermaLink="true">https://agentic-madness.com/en/posts/two-ways-to-make-an-agent-read-logs/</guid><description>Log Validate checks a log against expectations extracted from the code; Log Insight reconstructs system behavior from complete log chunks. How they differ and in what order to use them.</description><pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate><category>skill</category></item></channel></rss>