<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Model Context Protocol on Coderrob</title><link>https://coderrob.com/tags/model-context-protocol/</link><description>Recent content in Model Context Protocol on Coderrob</description><generator>Hugo</generator><language>en-US</language><lastBuildDate>Sat, 19 Sep 2026 10:00:00 -0500</lastBuildDate><atom:link href="https://coderrob.com/tags/model-context-protocol/index.xml" rel="self" type="application/rss+xml"/><item><title>Improving the Backstage MCP Server with MCP Kernel</title><link>https://coderrob.com/posts/improving-backstage-mcp-server-with-mcp-kernel/</link><pubDate>Sat, 19 Sep 2026 10:00:00 -0500</pubDate><guid>https://coderrob.com/posts/improving-backstage-mcp-server-with-mcp-kernel/</guid><description>&lt;p&gt;An MCP server for Backstage needs to preserve the meaning of a Catalog operation all the way from the tool call to the backend. A valid JSON response is only useful if the server asked the right question.&lt;/p&gt;&#10;&lt;p&gt;I&amp;rsquo;ve been updating my &lt;strong&gt;&lt;a href="https://github.com/Coderrob/backstage-mcp-server"&gt;Backstage MCP Server&lt;/a&gt;&lt;/strong&gt; around that requirement: improve the Catalog integration, make the runtime boundaries explicit, and verify the process an MCP client actually launches.&lt;/p&gt;&#10;&lt;p&gt;If you are new to the project, start with &lt;a href="https://coderrob.com/posts/introducing-backstage-mcp-server/"&gt;Introducing the Backstage MCP Server&lt;/a&gt; for what it does, example Catalog lookups, and client setup.&lt;/p&gt;</description></item><item><title>Creating and Publishing MCP Kernel</title><link>https://coderrob.com/posts/creating-and-publishing-mcp-kernel/</link><pubDate>Fri, 18 Sep 2026 10:00:00 -0500</pubDate><guid>https://coderrob.com/posts/creating-and-publishing-mcp-kernel/</guid><description>&lt;p&gt;An MCP server needs more than a collection of tool handlers. Once you add service clients, shared context, authorization rules, and cleanup, you have an application to maintain.&lt;/p&gt;&#10;&lt;p&gt;I created and published &lt;strong&gt;MCP Kernel&lt;/strong&gt; to handle that repeated application code. It gives TypeScript MCP servers reusable plugins, typed context, shared request policies, and lifecycle management so you can spend more time building the tools your application needs.&lt;/p&gt;&#10;&lt;p&gt;You can find it on &lt;a href="https://github.com/Coderrob/mcp-kernel"&gt;GitHub&lt;/a&gt; and &lt;a href="https://www.npmjs.com/package/@coderrob/mcp-kernel"&gt;npm&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Introducing the Backstage MCP Server</title><link>https://coderrob.com/posts/introducing-backstage-mcp-server/</link><pubDate>Thu, 17 Sep 2026 10:00:00 -0500</pubDate><guid>https://coderrob.com/posts/introducing-backstage-mcp-server/</guid><description>&lt;p&gt;If your team uses Backstage, your software catalog already contains information an AI assistant needs: service names, owners, APIs, and relationships between components. Giving the assistant access to that information can save you from copying catalog entries into a conversation every time you ask about a service.&lt;/p&gt;&#10;&lt;p&gt;I built &lt;strong&gt;&lt;a href="https://github.com/Coderrob/backstage-mcp-server"&gt;Backstage MCP Server&lt;/a&gt;&lt;/strong&gt; to connect an MCP-compatible assistant to the Backstage Software Catalog. It exposes tools for looking up entities, inspecting relationships, validating descriptors, and making authorized catalog changes.&lt;/p&gt;</description></item></channel></rss>