<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title>AI Agent Security - Tag - Jun Wen's Blog</title><link>https://junwen7.com/tags/ai-agent-security/</link><description>AI Agent Security - Tag - Jun Wen's Blog</description><generator>Hugo -- gohugo.io</generator><language>en</language><lastBuildDate>Wed, 27 May 2026 00:00:00 -0700</lastBuildDate><atom:link href="https://junwen7.com/tags/ai-agent-security/" rel="self" type="application/rss+xml"/><item><title>AI Agents in EDR Telemetry: Coverage and Blind Spots</title><link>https://junwen7.com/posts/ai_agents_edr_telemetry/</link><pubDate>Wed, 27 May 2026 00:00:00 -0700</pubDate><author><name>Jun Wen</name></author><guid>https://junwen7.com/posts/ai_agents_edr_telemetry/</guid><description>&lt;div class="featured-image">
&lt;img src="/images/ai_edr/cover.png" referrerpolicy="no-referrer">
&lt;/div>AI coding agents run with broad system privileges and generate a stream of EDR telemetry that is hard to interpret. This post maps each agent tool call to its host-level footprint — process, file, and network — across Claude Code and OpenCode, and explains where EDR visibility holds up and where it breaks down.</description></item></channel></rss>