<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Mcp on Security Unlocked</title><link>https://securityunlocked.com/tags/mcp/</link><description>Recent content in Mcp on Security Unlocked</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Tue, 28 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://securityunlocked.com/tags/mcp/index.xml" rel="self" type="application/rss+xml"/><item><title>Agentic Trust Debt: How 'Agent-Controlled Input' Became the New Buffer Overflow</title><link>https://securityunlocked.com/articles/agentic-trust-debt-how-agent-controlled-input-became-the-new-buffer-overflow/</link><pubDate>Tue, 28 Apr 2026 00:00:00 +0000</pubDate><guid>https://securityunlocked.com/articles/agentic-trust-debt-how-agent-controlled-input-became-the-new-buffer-overflow/</guid><description>Five AI agent frameworks disclosed the same vulnerability class in a single week, and the MCP SDK STDIO injection extended the pattern across four language ecosystems. The cluster reads like the buffer overflow era: a field-level conceptual gap in how agentic systems handle trust, not a string of individual implementation bugs.</description></item><item><title>AI Infrastructure Exploited Within 24 Hours of Disclosure</title><link>https://securityunlocked.com/weekly-intelligence/the-advisory-is-the-starting-gun/</link><pubDate>Mon, 27 Apr 2026 00:00:00 +0000</pubDate><guid>https://securityunlocked.com/weekly-intelligence/the-advisory-is-the-starting-gun/</guid><description>Four AI infrastructure platforms (Langflow, Marimo, LMDeploy, Flowise) were exploited within 24 hours of vulnerability disclosure last week. The patching window has collapsed to under one attacker shift.</description></item><item><title>The Protocol Is Doing Its Job</title><link>https://securityunlocked.com/weekly-intelligence/the-protocol-is-doing-its-job/</link><pubDate>Sun, 19 Apr 2026 00:00:00 +0000</pubDate><guid>https://securityunlocked.com/weekly-intelligence/the-protocol-is-doing-its-job/</guid><description>MCP&amp;rsquo;s trust architecture makes any exposed management interface a pre-authenticated command shell by design, not by accident, and two RCE vulnerabilities in the same week reveal a deployment curve that has outrun both audit methodology and detection playbooks.</description></item></channel></rss>