<?xml version="1.0" encoding="UTF-8" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>PentStark Research</title>
    <link>https://pentstark.com/blog</link>
    <description>CVE analyses, TTP walkthroughs, and methodology deep-dives from PentStark.</description>
    <language>en-us</language>
    <lastBuildDate>Fri, 24 Apr 2026 06:21:58 GMT</lastBuildDate>
    <atom:link href="https://pentstark.com/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>MCP servers are the new vendor risk: auditing the agent toolbox</title>
      <link>https://pentstark.com/blog/mcp-servers-are-the-new-vendor-risk</link>
      <guid isPermaLink="true">https://pentstark.com/blog/mcp-servers-are-the-new-vendor-risk</guid>
      <pubDate>Wed, 22 Apr 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark AI Research)</author>
      <category>AI Security</category>
      <description>Every MCP server you plug into your agent is a trust boundary. Here&apos;s how to audit them like the supply-chain risk they are.</description>
    </item>
    <item>
      <title>Passkeys in the enterprise: 18 months in, here&apos;s what breaks</title>
      <link>https://pentstark.com/blog/passkeys-enterprise-18-months-in</link>
      <guid isPermaLink="true">https://pentstark.com/blog/passkeys-enterprise-18-months-in</guid>
      <pubDate>Sat, 18 Apr 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Red Team)</author>
      <category>Identity</category>
      <description>Five recurring attack paths we&apos;ve found in enterprise passkey rollouts since they became table stakes.</description>
    </item>
    <item>
      <title>EU AI Act red teaming: what a pentester actually delivers</title>
      <link>https://pentstark.com/blog/eu-ai-act-red-teaming</link>
      <guid isPermaLink="true">https://pentstark.com/blog/eu-ai-act-red-teaming</guid>
      <pubDate>Wed, 15 Apr 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Compliance)</author>
      <category>Compliance</category>
      <description>Article 15 and 54a say you need cybersecurity and adversarial testing. Here&apos;s the engagement shape the first audit cycle accepted.</description>
    </item>
    <item>
      <title>AD CS ESC1 in the wild: the certificate template you forgot about</title>
      <link>https://pentstark.com/blog/ad-cs-esc1-misconfiguration</link>
      <guid isPermaLink="true">https://pentstark.com/blog/ad-cs-esc1-misconfiguration</guid>
      <pubDate>Fri, 10 Apr 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Red Team)</author>
      <category>Active Directory</category>
      <description>How a single ManageCA + supplied-subject template got us to domain admin in 22 minutes.</description>
    </item>
    <item>
      <title>Your Kubernetes ingress is the new DMZ</title>
      <link>https://pentstark.com/blog/kubernetes-ingress-is-the-new-dmz</link>
      <guid isPermaLink="true">https://pentstark.com/blog/kubernetes-ingress-is-the-new-dmz</guid>
      <pubDate>Fri, 03 Apr 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Cloud)</author>
      <category>Cloud Security</category>
      <description>Ingress controllers parse untrusted input, reach every service, and run privileged. They are a DMZ, not a detail.</description>
    </item>
    <item>
      <title>LLM tool-use escalation: from prompt injection to arbitrary code</title>
      <link>https://pentstark.com/blog/llm-tool-use-escalation</link>
      <guid isPermaLink="true">https://pentstark.com/blog/llm-tool-use-escalation</guid>
      <pubDate>Sat, 28 Mar 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark AI Research)</author>
      <category>AI Security</category>
      <description>A case study on bridging a benign prompt-injection primitive into full agent compromise.</description>
    </item>
    <item>
      <title>DORA ICT third-party testing: lessons from the first audit cycle</title>
      <link>https://pentstark.com/blog/dora-ict-testing-first-cycle-lessons</link>
      <guid isPermaLink="true">https://pentstark.com/blog/dora-ict-testing-first-cycle-lessons</guid>
      <pubDate>Fri, 20 Mar 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Compliance)</author>
      <category>Compliance</category>
      <description>What we learned running threat-led penetration tests under DORA Articles 26–27 across the first wave of 2026 audits.</description>
    </item>
    <item>
      <title>BOLA across microservices: when authorization is someone else&apos;s job</title>
      <link>https://pentstark.com/blog/bola-across-microservices</link>
      <guid isPermaLink="true">https://pentstark.com/blog/bola-across-microservices</guid>
      <pubDate>Thu, 12 Mar 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark PTaaS)</author>
      <category>AppSec</category>
      <description>Why BOLA is the most under-tested flaw class in modern SaaS — and how we find it.</description>
    </item>
    <item>
      <title>Lockfile injection: the npm supply-chain attack your SCA missed</title>
      <link>https://pentstark.com/blog/lockfile-injection-supply-chain</link>
      <guid isPermaLink="true">https://pentstark.com/blog/lockfile-injection-supply-chain</guid>
      <pubDate>Thu, 05 Mar 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Product Security)</author>
      <category>AppSec</category>
      <description>Pinning your lockfile is necessary but not sufficient. Here&apos;s the pattern attackers are using to slip past SCA tools.</description>
    </item>
    <item>
      <title>Ransomware-ready posture: the assume-breach scenario every board should fund</title>
      <link>https://pentstark.com/blog/ransomware-ready-posture-check</link>
      <guid isPermaLink="true">https://pentstark.com/blog/ransomware-ready-posture-check</guid>
      <pubDate>Fri, 20 Feb 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Red Team)</author>
      <category>Red Team</category>
      <description>A red-team scoping guide for boards asking &apos;are we ransomware-ready?&apos;</description>
    </item>
    <item>
      <title>SOC 2 without slowing down your release train</title>
      <link>https://pentstark.com/blog/soc2-without-slowing-down</link>
      <guid isPermaLink="true">https://pentstark.com/blog/soc2-without-slowing-down</guid>
      <pubDate>Wed, 04 Feb 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Compliance)</author>
      <category>Compliance</category>
      <description>Patterns for embedding evidence collection into your engineering workflow.</description>
    </item>
    <item>
      <title>Shift-left that actually ships: patterns from ten SSDLC rollouts</title>
      <link>https://pentstark.com/blog/ssdlc-shift-left-patterns</link>
      <guid isPermaLink="true">https://pentstark.com/blog/ssdlc-shift-left-patterns</guid>
      <pubDate>Thu, 22 Jan 2026 00:00:00 GMT</pubDate>
      <author>noreply@pentstark.com (PentStark Product Security)</author>
      <category>Product Security</category>
      <description>The three patterns that make product-security programs work — and the three that quietly kill them.</description>
    </item>
  </channel>
</rss>