<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Security Automation | Gelei Deng</title><link>https://geleideng.github.io/tags/security-automation/</link><atom:link href="https://geleideng.github.io/tags/security-automation/index.xml" rel="self" type="application/rss+xml"/><description>Security Automation</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Thu, 19 Feb 2026 00:00:00 +0000</lastBuildDate><image><url>https://geleideng.github.io/media/icon_hu7729264130191091259.png</url><title>Security Automation</title><link>https://geleideng.github.io/tags/security-automation/</link></image><item><title>Excalibur</title><link>https://geleideng.github.io/project/excalibur/</link><pubDate>Thu, 19 Feb 2026 00:00:00 +0000</pubDate><guid>https://geleideng.github.io/project/excalibur/</guid><description>&lt;p>Excalibur is a difficulty-aware LLM agent design for real-world penetration testing.&lt;/p>
&lt;p>It couples typed tooling, retrieval-augmented security knowledge, task difficulty assessment, and evidence-guided attack tree search to reduce planning failures in multi-step penetration testing tasks.&lt;/p></description></item><item><title>What Makes a Good LLM Agent for Real-world Penetration Testing?</title><link>https://geleideng.github.io/publication/llm-agent-pentest/</link><pubDate>Thu, 19 Feb 2026 00:00:00 +0000</pubDate><guid>https://geleideng.github.io/publication/llm-agent-pentest/</guid><description>&lt;p>This paper examines why LLM penetration testing systems succeed or fail in real-world settings, then proposes Excalibur to improve task selection and attack-chain planning through difficulty-aware reasoning.&lt;/p></description></item></channel></rss>