<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Margalho — Journal</title><description>Designing resilient software harnesses, multi-agent orchestration loops, and AI systems that actually hold up in production.</description><link>https://margalho.xyz/</link><language>en-us</language><item><title>Why the Harness Matters More Than the Model</title><link>https://margalho.xyz/blog/why-the-harness-matters/</link><guid isPermaLink="true">https://margalho.xyz/blog/why-the-harness-matters/</guid><description>From the tiny alien inside Rosenberg&apos;s head to Russian nesting dolls, multi-model subagents, and why production reliability is engineered in the harness rather than inside the weights.</description><pubDate>Thu, 12 Nov 2026 00:00:00 GMT</pubDate><category>en</category><category>essay</category><category>harness</category><category>ai-engineering</category><category>agent-orchestration</category><category>systems</category><author>hello@margalho.xyz (Margalho)</author></item></channel></rss>