<?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 Guillaume Bogard</title><link>/tags/ai/</link><description>Recent content in Ai on Guillaume Bogard</description><generator>Hugo</generator><language>en</language><lastBuildDate>Thu, 02 Jul 2026 18:22:14 +0200</lastBuildDate><atom:link href="/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Declarative, harness-agnostic agent skills with Nix</title><link>/posts/rigup/</link><pubDate>Thu, 02 Jul 2026 00:00:00 +0000</pubDate><guid>/posts/rigup/</guid><description>&lt;p&gt;It&amp;rsquo;s been a while since I&amp;rsquo;ve posted anything on this blog. Today I&amp;rsquo;d like to promote a tool that my friend &lt;a href="https://github.com/YPares"&gt;@YPares&lt;/a&gt;
has created, and which I&amp;rsquo;ve been daily driving for several months.&lt;/p&gt;
&lt;p&gt;&lt;a href="https://github.com/YPares/rigup.nix"&gt;Rigup&lt;/a&gt; is a Nix-based tool that allows you to bundle together instructions, knowledge, and executable programs
to form skills for your AI agent. These skills, called &lt;em&gt;riglets&lt;/em&gt; in Rigup jargon, are not bound to a specific AI harness, and don&amp;rsquo;t require that you
pollute your workspace with globally-installed programs.&lt;/p&gt;</description></item></channel></rss>