<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Agent Architecture on Andreas Nissen</title><link>https://andreasnissen.dev/tags/agent-architecture/</link><description>Recent content in Agent Architecture on Andreas Nissen</description><generator>Hugo</generator><language>en-us</language><copyright>© Andreas Nissen</copyright><lastBuildDate>Sun, 30 Aug 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://andreasnissen.dev/tags/agent-architecture/index.xml" rel="self" type="application/rss+xml"/><item><title>Context and Control Are Different Layers</title><link>https://andreasnissen.dev/writing/context-and-control/</link><pubDate>Sun, 30 Aug 2026 00:00:00 +0000</pubDate><guid>https://andreasnissen.dev/writing/context-and-control/</guid><description>&lt;p&gt;Enterprise AI agents need context and control.&lt;/p&gt;&#10;&lt;p&gt;Context without control is unsafe. Control without context is blind.&lt;/p&gt;&#10;&lt;p&gt;They belong in the same architecture, and they should not collapse into the same layer. A missing relationship degrades an answer. A missing authorization check can trigger an action the user never permitted. Those failures have different consequences, owners, and tests.&lt;/p&gt;&#10;&lt;h2 id="context-answers-what-the-agent-should-know"&gt;Context answers what the agent should know&lt;/h2&gt;&#10;&lt;p&gt;Context is the evidence available for reasoning. It can include documents, structured records, tool results, conversation state, identity attributes, and the relationships between them.&lt;/p&gt;</description></item></channel></rss>