<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Jeet Jain</title>
    <description>The latest articles on DEV Community by Jeet Jain (@freeguy21).</description>
    <link>https://dev.to/freeguy21</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3935131%2Fc71946b3-9a7b-47d6-8db2-ed1ddd2756f4.png</url>
      <title>DEV Community: Jeet Jain</title>
      <link>https://dev.to/freeguy21</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/freeguy21"/>
    <language>en</language>
    <item>
      <title>MCP Security is Broken — So I Built a Scanner</title>
      <dc:creator>Jeet Jain</dc:creator>
      <pubDate>Sat, 16 May 2026 16:10:16 +0000</pubDate>
      <link>https://dev.to/freeguy21/mcp-security-is-broken-so-i-built-a-scanner-30ak</link>
      <guid>https://dev.to/freeguy21/mcp-security-is-broken-so-i-built-a-scanner-30ak</guid>
      <description>&lt;p&gt;MCP (Model Context Protocol) is the new standard for connecting AI agents to tools. The security is a disaster right now.&lt;/p&gt;

&lt;p&gt;BlueRock Security scanned 7,000+ live MCP servers 36.7% were vulnerable to SSRF. Hundreds had zero authentication or encryption.&lt;/p&gt;

&lt;p&gt;So I built AgentWarden — a CLI tool that scans MCP servers for real vulnerabilities:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Unauthenticated endpoints&lt;/li&gt;
&lt;li&gt;SSRF via tool parameters
&lt;/li&gt;
&lt;li&gt;Prompt injection in tool descriptions (tool poisoning)&lt;/li&gt;
&lt;li&gt;Path traversal in file tools&lt;/li&gt;
&lt;li&gt;Tool shadowing and typosquatting&lt;/li&gt;
&lt;li&gt;Missing rate limiting&lt;/li&gt;
&lt;li&gt;TLS misconfigurations&lt;/li&gt;
&lt;li&gt;Sensitive data exposure&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Usage:&lt;br&gt;
agentwarden scan &lt;a href="https://your-mcp-server.com" rel="noopener noreferrer"&gt;https://your-mcp-server.com&lt;/a&gt; -v&lt;br&gt;
agentwarden scan &lt;a href="https://your-mcp-server.com" rel="noopener noreferrer"&gt;https://your-mcp-server.com&lt;/a&gt; -o report.html&lt;/p&gt;

&lt;p&gt;GitHub: &lt;a href="https://github.com/Agent-Warden/Agent-Warden" rel="noopener noreferrer"&gt;https://github.com/Agent-Warden/Agent-Warden&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Scanned several live MCP servers already. Would love feedback/contribution from anyone in the security space building on MCP.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>cybersecurity</category>
      <category>mcp</category>
      <category>opensource</category>
    </item>
  </channel>
</rss>
