<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai on Munatsi Ziumbe</title><link>https://munatsi.ziumbe.com/tags/ai/</link><description>Recent content in Ai on Munatsi Ziumbe</description><generator>Hugo</generator><language>en-gb</language><lastBuildDate>Tue, 17 Mar 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://munatsi.ziumbe.com/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>The SQLite Principles for Agent-Ready APIs</title><link>https://munatsi.ziumbe.com/posts/sqlite-principles-for-agent-tools/</link><pubDate>Tue, 17 Mar 2026 00:00:00 +0000</pubDate><guid>https://munatsi.ziumbe.com/posts/sqlite-principles-for-agent-tools/</guid><description>&lt;p>&lt;em>This manifesto emerged from a conversation I had with an AI about what makes tools work well for agents. We started from SQLite&amp;rsquo;s design philosophy and worked out what it teaches about building tools that AI can actually use reliably.&lt;/em>&lt;/p>
&lt;hr>
&lt;blockquote>
&lt;p>What embedded database design teaches us about building tools that AI can actually use.&lt;/p>&lt;/blockquote>
&lt;h2 id="preamble">Preamble&lt;/h2>
&lt;p>SQLite is one of the most widely deployed pieces of software in existence. Its secret is not the database engine — it is the API. Eight functions. Two files. No ceremony. A developer can read the entire documentation and hold a complete mental model in their head by the end of an afternoon.&lt;/p></description></item></channel></rss>