<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by reza gholizade on Medium]]></title>
        <description><![CDATA[Stories by reza gholizade on Medium]]></description>
        <link>https://medium.com/@gholizade.net?source=rss-2954acd69ef9------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*LiQGf5DActfcJdhW.</url>
            <title>Stories by reza gholizade on Medium</title>
            <link>https://medium.com/@gholizade.net?source=rss-2954acd69ef9------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 23:33:28 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@gholizade.net/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[️ Helm Support for k8s-mcp-server + Updated n8n Workflow]]></title>
            <link>https://medium.com/@gholizade.net/%EF%B8%8F-helm-support-for-k8s-mcp-server-updated-n8n-workflow-e277c83c3236?source=rss-2954acd69ef9------2</link>
            <guid isPermaLink="false">https://medium.com/p/e277c83c3236</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[model-context-protocol]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[k8s]]></category>
            <category><![CDATA[mcp-server]]></category>
            <dc:creator><![CDATA[reza gholizade]]></dc:creator>
            <pubDate>Wed, 16 Jul 2025 12:00:44 GMT</pubDate>
            <atom:updated>2025-07-19T13:43:37.469Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7dkCUTRg-fSsw2Jk0opHEQ.png" /></figure><p>In my <a href="https://medium.com/@gholizade.net/introducing-k8s-mcp-server-a-secure-lightweight-gateway-between-ai-agents-and-kubernetes-e9cdcbe09d70">previous post</a>, I introduced <a href="https://github.com/reza-gholizade/k8s-mcp-server">k8s-mcp-server</a> — a secure, lightweight gateway between AI agents and Kubernetes clusters. The idea was to enable LLM-powered workflows and assistants to interact with K8s safely, without exposing the full cluster control.</p><p>Now, I’m excited to announce a major update to the project:</p><p>✅ <strong>Helm Tools Support</strong><br>✅ <strong>Updated n8n Workflow with Helm Integration</strong></p><p>Let’s dive into what’s new 👇</p><p>Now, I’m excited to announce a major update to the project:</p><p>✅ <strong>Helm Tools Support</strong><br> ✅ <strong>Updated n8n Workflow with Helm Integration</strong></p><p>Let’s dive into what’s new 👇</p><h3>🎯 Why Helm?</h3><p><a href="https://helm.sh/">Helm</a> is the most widely adopted package manager for Kubernetes. It simplifies deployment, upgrades, and templating of K8s resources. Integrating Helm into the MCP server enables more powerful workflows for both human operators and LLM agents.</p><h3>🚀 New Helm Tools (v0.2.0+)</h3><p>The MCP server now supports Helm tooling as secure API endpoints, giving your workflows and agents the ability to:</p><ul><li>List Helm releases in a namespace</li><li>Get details of a specific release</li><li>Fetch and render templates (without applying them)</li><li>Install or upgrade charts</li><li>Uninstall releases</li><li>Show current release values</li><li>Helm rollback support</li></ul><p>All with permission-aware, controlled execution.</p><p>🧪 Example commands agents can now issue:</p><pre>&quot;List all Helm releases in the production namespace&quot;<br>&quot;Upgrade the API-Gateway to version 2.0.1.&quot;<br>&quot;Show current values of the nginx release. &quot;For full API reference, visit the GitHub repo.</pre><h3>🔄 Updated n8n Workflow with Helm Integration</h3><p>We’ve also updated the official <a href="https://n8n.io/workflows/4023-conversational-kubernetes-management-with-gpt-4o-and-mcp-integration/">n8n</a> workflow to include support for Helm-based tools. This means your automated workflows and agents can now:</p><ul><li>Trigger Helm actions via chat or HTTP</li><li>Log Helm events for auditability</li><li>React to alerts and incidents (e.g., rollback after a failed release)</li></ul><p>🔧 Example use cases:</p><ul><li>A Telegram chatbot that rolls back a release after a Prometheus alert</li><li>A Slack command to list Helm releases in staging</li><li>Auto-previewing Helm templates before applying changes</li></ul><p>You can find the updated workflow in this <a href="https://n8n.io/workflows/4023-conversational-kubernetes-management-with-gpt-4o-and-mcp-integration/">Link</a>.</p><h3>🔐 Security Considerations</h3><p>As with the rest of the MCP server, all Helm actions pass through the secure proxy, which supports optional token-based authentication, namespace scoping, and execution logging. This ensures that even powerful tools like helm upgrade don’t become dangerous in the wrong hands.</p><h3>📦 Try It Out</h3><p>You can get started right now:</p><ul><li>🧠 GitHub: <a href="https://github.com/reza-gholizade/k8s-mcp-server">github.com/reza-gholizade/k8s-mcp-server</a></li><li>🐳 Docker: ginnux/k8s-mcp-server:latest</li></ul><p>Simply pull the latest version and enable the Helm toolset in your config.</p><h3>🙌 Feedback &amp; Contributions Welcome</h3><p>If you’re building LLM-powered tools, CI/CD bots, or Kubernetes automation workflows, give k8s-mcp-server a try. I’d love your feedback, issues, and contributions.</p><p>Star the repo if you find it useful ⭐<br> Cheers!<br> <br><a href="https://www.linkedin.com/in/reza-gholizade/">Reza Gholizade</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e277c83c3236" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ Introducing k8s-mcp-server: A Secure, Lightweight Gateway Between AI Agents and Kubernetes]]></title>
            <link>https://medium.com/@gholizade.net/introducing-k8s-mcp-server-a-secure-lightweight-gateway-between-ai-agents-and-kubernetes-e9cdcbe09d70?source=rss-2954acd69ef9------2</link>
            <guid isPermaLink="false">https://medium.com/p/e9cdcbe09d70</guid>
            <category><![CDATA[mcp-server]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[go]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[docker]]></category>
            <dc:creator><![CDATA[reza gholizade]]></dc:creator>
            <pubDate>Mon, 19 May 2025 14:41:54 GMT</pubDate>
            <atom:updated>2025-05-30T13:36:43.228Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><em>Simplifying Kubernetes interactions for automation, observability, and AI agent integration — built in Go, with multi-cluster support and SSE-powered APIs.</em></blockquote><h3>🌐 Overview</h3><p>In the world of platform engineering and intelligent automation, accessing and interacting with Kubernetes clusters securely and efficiently is essential. Whether you’re building an AI-powered DevOps assistant, an internal developer platform, or a bot that helps your SRE team troubleshoot faster — having a clean, safe, and powerful interface to Kubernetes is critical.</p><p>Here’s the n8n workflow template:</p><p><a href="https://n8n.io/workflows/4023-conversational-kubernetes-management-with-gpt-4o-and-mcp-integration/">https://n8n.io/workflows/4023-conversational-kubernetes-management-with-gpt-4o-and-mcp-integration/</a></p><p>🧠 <strong>Real Example: n8n + k8s-mcp-server + LLM</strong></p><p>In the screenshot below, you can see an actual <strong>n8n workflow</strong> where I’ve integrated k8s-mcp-server with an LLM-powered chatbot. I simply told n8n:</p><blockquote>“Summarize the last 10 lines of pod chatbot in namespace ai”</blockquote><p>The LLM parses the instruction, uses k8s-mcp-server to fetch logs via the Kubernetes API, and returns a concise summary — all automated in real-time. This demonstrates how easily AI agents can interact with Kubernetes through clean, safe abstractions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cII65eel2f8yQzaexdVtJw.png" /></figure><p>That’s why I built <a href="https://github.com/reza-gholizade/k8s-mcp-server"><strong>k8s-mcp-server</strong></a> — a lightweight and secure <strong>Model Context Protocol (MCP) server</strong> that exposes real-time, event-driven endpoints for Kubernetes resource interaction, designed for multi-cluster environments and built to play nicely with AI agent ecosystems like <strong>LangChain</strong>, <strong>n8n</strong>, <strong>OpenAI</strong>, <strong>Claude</strong>, and <strong>Cursor</strong>.</p><h3>🧠 What is k8s-mcp-server?</h3><p>k8s-mcp-server is a Go-based backend server that acts as a <strong>safe and extensible bridge</strong> between Kubernetes clusters and external systems. It allows users or agents to <strong>query, list, describe, and even create</strong> Kubernetes resources via a clean HTTP/SSE API interface.</p><p>It’s ideal for building:</p><ul><li>🤖 LLM-based Kubernetes assistants</li><li>🔧 Internal tools for DevOps/SREs</li><li>📊 Observability dashboards</li><li>🧩 Event-driven automation pipelines</li></ul><h3>🚀 Key Features</h3><h3>📦 SSE APIs for Real-time Interaction</h3><p>The server provides <strong>Server-Sent Events (SSE)</strong> endpoints for streaming resource data in real-time — perfect for AI assistants, dashboards, or chatbots.</p><h3>🔒 Safe by Default</h3><p>Critical destructive commands like delete, exec, and port-forward are <strong>blocked</strong> by design — reducing risk in automated systems.</p><h3>🌐 Multi-cluster Ready</h3><p>Deploy multiple instances of the MCP server across your Kubernetes clusters and route them as needed. Ideal for hybrid cloud and federated environments.</p><h3>🤖 AI &amp; Agent Friendly</h3><p>Integrates easily with modern agent stacks such as:</p><ul><li><a href="https://www.langchain.com/">LangChain</a></li><li><a href="https://n8n.io/">n8n.io</a></li><li><a href="https://openai.com/">OpenAI GPT Agents</a></li><li><a href="https://www.anthropic.com/">Claude by Anthropic</a></li><li><a href="https://cursor.sh/">Cursor.sh</a></li></ul><h3>⚙️ Built in Go</h3><p>The entire system is written in Go, ensuring high performance, low memory usage, and fast startup.</p><h3>📥 How to Use</h3><p>Pull the latest image:</p><pre>docker pull ginnux/k8s-mcp-server</pre><p>Or run it in Kubernetes:</p><pre>apiVersion: apps/v1<br>kind: Deployment<br>metadata:<br>  name: mcp-server<br>spec:<br>  replicas: 1<br>  selector:<br>    matchLabels:<br>      app: mcp-server<br>  template:<br>    metadata:<br>      labels:<br>        app: mcp-server<br>    spec:<br>      containers:<br>        - name: mcp-server<br>          image: ginnux/k8s-mcp-server:latest<br>          ports:<br>            - containerPort: 8080</pre><h3>🔗 GitHub Repository</h3><p>Explore the source, contribute, or file an issue:</p><p>👉 <a href="https://github.com/reza-gholizade/k8s-mcp-server">https://github.com/reza-gholizade/k8s-mcp-server</a></p><h3>🎯 Final Thoughts</h3><p>With the rise of AI-native developer experiences and multi-cluster architectures, tools like k8s-mcp-server help bridge the gap between low-level infrastructure and high-level automation.</p><p>Whether you’re building a GPT-powered Kubernetes CLI, connecting agents to your clusters, or simply looking for a safe abstraction layer , give k8s-mcp-server a try.</p><p>✉️ <strong>Feedback and Contributions Welcome!</strong><br> If you have ideas, questions, or suggestions, feel free to open an issue or contribute via GitHub.</p><p>Happy Hacking 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e9cdcbe09d70" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Custom Kubernetes Control Plane with k8s-mcp-server]]></title>
            <link>https://medium.com/@gholizade.net/building-a-custom-kubernetes-control-plane-with-k8s-mcp-server-dadd0db97439?source=rss-2954acd69ef9------2</link>
            <guid isPermaLink="false">https://medium.com/p/dadd0db97439</guid>
            <category><![CDATA[mcp-server]]></category>
            <category><![CDATA[k8s]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[reza gholizade]]></dc:creator>
            <pubDate>Sun, 27 Apr 2025 19:33:41 GMT</pubDate>
            <atom:updated>2025-04-27T19:34:26.959Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8wEFxT2lRzWRLJVqb7mxOQ.png" /></figure><p>In the ever-evolving cloud-native landscape, Kubernetes has become the de-facto standard for container orchestration. Yet, interacting with Kubernetes resources consistently across different clusters and contexts can sometimes be complex.<br> To address this, I built <strong>k8s-mcp-server</strong> — a lightweight Model Control Plane (MCP) server designed to offer a standardized, flexible interface for Kubernetes cluster operations.</p><p>In this post, I’ll walk you through what <strong>k8s-mcp-server</strong> is, why I built it, and how you can start using it today.</p><h3>What is k8s-mcp-server?</h3><p><strong>k8s-mcp-server</strong> is a custom Kubernetes MCP (Model Control Plane) server that provides a simple, standardized way to interact with Kubernetes resources.<br> It allows tools or clients to list, inspect, describe, and even fetch logs from Kubernetes resources using a <strong>JSON-RPC</strong> based protocol over standard input/output (stdin/stdout).</p><p>In short: it’s like a programmable, lightweight kubectl alternative for automated systems and custom tooling.</p><h3>Key Features</h3><ul><li><strong>API Resource Discovery</strong>: Easily discover all API resources available in your cluster.</li><li><strong>Flexible Resource Listing</strong>: List any Kubernetes resources, filtered by namespace or labels.</li><li><strong>Resource Inspection</strong>: Retrieve detailed information about any resource instance.</li><li><strong>Resource Description</strong>: Get comprehensive, kubectl describe-style descriptions for resources.</li><li><strong>Pod Log Retrieval</strong>: Fetch logs from running pods directly through the protocol.</li><li><strong>Standardized Communication</strong>: All operations follow a consistent MCP (Model Control Plane) protocol.</li><li><strong>Multi-Cluster Friendly</strong>: Supports working with different Kubernetes contexts and scopes.</li></ul><h3>Why I Built It</h3><p>While kubectl is a powerful CLI tool, integrating it into custom automation workflows can be tedious, especially when a standardized communication protocol is needed.<br> I wanted a lightweight server that:</p><ul><li><strong>Speaks a predictable protocol</strong> (MCP, JSON-RPC).</li><li><strong>Is easy to embed</strong> inside larger systems (agents, operators, dashboards).</li><li><strong>Doesn’t need heavy client libraries</strong> — just send simple messages over stdin/stdout.</li></ul><p>Thus, <strong>k8s-mcp-server</strong> was born.</p><h3>Prerequisites</h3><p>To run <strong>k8s-mcp-server</strong>, you’ll need:</p><ul><li><strong>Go 1.20+</strong> is installed on your machine.</li><li>Access to a <strong>Kubernetes cluster</strong>.</li><li>A properly configured <strong>kubectl</strong> (kubeconfig).</li></ul><h3>Installation</h3><p>Clone the repository:</p><pre>git clone https://github.com/reza-gholizade/k8s-mcp-server.git<br>cd k8s-mcp-server</pre><p>Download dependencies:</p><pre>go mod download</pre><p>Build the server:</p><pre>go build -o k8s-mcp-server main.go</pre><h3>Usage</h3><p>Start the server:</p><pre>./k8s-mcp-server</pre><p>The server listens for <strong>MCP messages</strong> on stdin and responds on stdout.</p><p>You can now interact with your Kubernetes cluster by sending JSON-RPC messages.<br> Here are some examples:</p><h3>1. Discover Available API Resources</h3><pre>{<br>  &quot;jsonrpc&quot;: &quot;2.0&quot;,<br>  &quot;id&quot;: 1,<br>  &quot;method&quot;: &quot;getAPIResources&quot;,<br>  &quot;params&quot;: {<br>    &quot;arguments&quot;: {<br>      &quot;includeNamespaceScoped&quot;: true,<br>      &quot;includeClusterScoped&quot;: true<br>    }<br>  }<br>}</pre><h3>2. List Specific Resources</h3><pre>{<br>  &quot;jsonrpc&quot;: &quot;2.0&quot;,<br>  &quot;id&quot;: 2,<br>  &quot;method&quot;: &quot;listResources&quot;,<br>  &quot;params&quot;: {<br>    &quot;arguments&quot;: {<br>      &quot;Kind&quot;: &quot;Pod&quot;,<br>      &quot;namespace&quot;: &quot;default&quot;,<br>      &quot;labelSelector&quot;: &quot;app=nginx&quot;<br>    }<br>  }<br>}</pre><h3>3. Get Details of a Specific Resource</h3><pre>{<br>  &quot;jsonrpc&quot;: &quot;2.0&quot;,<br>  &quot;id&quot;: 3,<br>  &quot;method&quot;: &quot;getResource&quot;,<br>  &quot;params&quot;: {<br>    &quot;arguments&quot;: {<br>      &quot;kind&quot;: &quot;Pod&quot;,<br>      &quot;name&quot;: &quot;nginx-pod&quot;,<br>      &quot;namespace&quot;: &quot;default&quot;<br>    }<br>  }<br>}</pre><h3>4. Describe a Resource</h3><pre>{<br>  &quot;jsonrpc&quot;: &quot;2.0&quot;,<br>  &quot;id&quot;: 4,<br>  &quot;method&quot;: &quot;describeResource&quot;,<br>  &quot;params&quot;: {<br>    &quot;arguments&quot;: {<br>      &quot;Kind&quot;: &quot;Deployment&quot;,<br>      &quot;name&quot;: &quot;my-app&quot;,<br>      &quot;namespace&quot;: &quot;production&quot;<br>    }<br>  }<br>}</pre><h3>5. Retrieve Pod Logs</h3><p>You can also fetch logs from a specific Pod:</p><pre>{<br>  &quot;jsonrpc&quot;: &quot;2.0&quot;,<br>  &quot;id&quot;: 5,<br>  &quot;method&quot;: &quot;getPodLogs&quot;,<br>  &quot;params&quot;: {<br>    &quot;arguments&quot;: {<br>      &quot;name&quot;: &quot;nginx-pod&quot;,<br>      &quot;namespace&quot;: &quot;default&quot;,<br>      &quot;container&quot;: &quot;nginx&quot;,<br>      &quot;tailLines&quot;: 100<br>    }<br>  }<br>}</pre><p><strong>Parameters include</strong>:</p><ul><li>name: Pod name (required)</li><li>namespace: Namespace of the Pod (required)</li><li>container: Specific container name (optional)</li><li>tailLines: Number of recent log lines to fetch (optional)</li></ul><p>This gives you an easy way to fetch pod logs without needing external tools.</p><h3>Final Thoughts</h3><p><strong>K8s-mcp-server</strong> is built to make Kubernetes interaction <strong>programmable</strong>, <strong>lightweight</strong>, and <strong>standardized</strong>.<br> Whether you’re developing an automation system, a custom Kubernetes dashboard, or an internal CLI tool, <strong>k8s-mcp-server</strong> gives you the building blocks to work with clusters cleanly and predictably.</p><p>I’m excited to continue improving it, adding authentication support, streaming, and multi-cluster management capabilities in the future.</p><p>If you’re interested, feel free to check out the source code on <a href="https://github.com/reza-gholizade/k8s-mcp-server">GitHub</a> and contribute!</p><p><strong>Thanks for reading!</strong><br> Follow me for more updates on Kubernetes, Go development, and cloud-native tools.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dadd0db97439" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>