<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Context on Aguilera Engineering</title><link>https://aguilera.ee/blog/tags/context/</link><description>Recent content in Context on Aguilera Engineering</description><generator>Hugo -- gohugo.io</generator><language>en-US</language><managingEditor>eduardo@aguilera.ee (Eduardo Aguilera)</managingEditor><webMaster>eduardo@aguilera.ee (Eduardo Aguilera)</webMaster><copyright>Eduardo Aguilera</copyright><lastBuildDate>Thu, 05 Mar 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://aguilera.ee/blog/tags/context/index.xml" rel="self" type="application/rss+xml"/><item><title>Skills over MCP servers</title><link>https://aguilera.ee/blog/skills-vs-mcp/</link><pubDate>Thu, 05 Mar 2026 00:00:00 +0000</pubDate><author>eduardo@aguilera.ee (Eduardo Aguilera)</author><guid>https://aguilera.ee/blog/skills-vs-mcp/</guid><description>&lt;p&gt;I use Skills instead of MCP servers.&lt;/p&gt;
&lt;p&gt;MCP servers port existing APIs into something agents can use. They also eat your
context window with verbose definitions.&lt;/p&gt;
&lt;p&gt;Skills are small markdown files with instructions for using CLI tools. They load
into context only when your question matches the skill&amp;rsquo;s description.&lt;/p&gt;
&lt;p&gt;I tested reading my reminders list through Claude Code. The Skill used 40
tokens. The MCP server used 749. Multiply that across the dozens of tools a power
user touches and Skills scale better.&lt;/p&gt;</description><content:encoded><![CDATA[<p>I use Skills instead of MCP servers.</p>
<p>MCP servers port existing APIs into something agents can use. They also eat your
context window with verbose definitions.</p>
<p>Skills are small markdown files with instructions for using CLI tools. They load
into context only when your question matches the skill&rsquo;s description.</p>
<p>I tested reading my reminders list through Claude Code. The Skill used 40
tokens. The MCP server used 749. Multiply that across the dozens of tools a power
user touches and Skills scale better.</p>
<p>Some of your MCP servers are a markdown file waiting to happen.</p>
]]></content:encoded></item></channel></rss>