# Agentic Kubernetes CLI > Agentic AI system that transforms natural language into kubectl commands via Strands Agent and MCP server integration. - **Type:** MCP server - **Install:** `agentstack add mcp-tarique-b-devops-agentic-kubernetes-cli` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [Tarique-B-DevOps](https://agentstack.voostack.com/s/tarique-b-devops) - **Installs:** 0 - **Category:** [Cloud & Infrastructure](https://agentstack.voostack.com/c/cloud-infrastructure) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [Tarique-B-DevOps](https://github.com/Tarique-B-DevOps) - **Source:** https://github.com/Tarique-B-DevOps/Agentic-Kubernetes-CLI ## Install ```sh agentstack add mcp-tarique-b-devops-agentic-kubernetes-cli ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Agentic Kubernetes CLI [](https://www.python.org/downloads/) [](https://kubernetes.io/) [](https://aws.amazon.com/bedrock/) [](LICENSE) > **An intelligent Agentic Kubernetes CLI assistant** that integrates your **local MCP server** with **AWS Bedrock**, enabling natural language control over **kubectl commands**, **cluster management**, and **real-time Kubernetes insights**. --- ## 🚀 Features ### 🗣️ Natural Language Kubernetes Operations - Conversational `kubectl` interface — execute commands in plain English - AI-powered command generation and validation - Context-aware management of clusters and namespaces - Real-time monitoring: pods, logs, resources, and events ### ⚙️ Advanced Capabilities - ✅ **Resource Management** — List, describe, and manage Kubernetes resources - 📊 **Troubleshooting Assistant** — Analyze logs and diagnose issues - 🔄 **Context Switching** — Move between clusters easily - ⚡ **Interactive CLI** — Friendly prompt-driven interface --- ## 🎥 Demo Below is an example interaction between the **User** and the **Agentic Kubernetes CLI** demonstrating real-time troubleshooting and remediation. --- **🧑‍💻 User:** list all pods in the cluster **🤖 Agent:** There are no running pods, but one pod `nginx` is showing an issue (`ImagePullBackOff`). **🧑‍💻 User:** analyze the issue with `nginx` pod **🤖 Agent:** Issue identified — the container image tag `nginx:lt` is incorrect and not found in the registry. **🧑‍💻 User:** fix the image tag issue and make the pod running **🤖 Agent:** Updated the image tag to `nginx:latest`, redeployed the pod, and verified that it is now **Running** successfully. --- Your browser does not support the video tag. [🎥 Watch Full Demo Video](https://github.com/user-attachments/assets/c30c4288-793c-4554-98ac-185075a621f0) --- ## 📚 Table of Contents - [Prerequisites](#-prerequisites) - [Installation](#-installation) - [Configuration](#-configuration) - [Usage](#-usage) - [Examples](#-examples) - [Architecture](#-architecture) - [MCP Tools](#-mcp-tools) --- ## 🔧 Prerequisites ### Required Software - [Python 3.12+](https://www.python.org/downloads/) - [kubectl](https://kubernetes.io/docs/tasks/tools/) - Access to a running **Kubernetes cluster** - **AWS account** with Bedrock API access ### AWS Bedrock Setup 1. Enable Claude models in the AWS Bedrock console 2. Configure AWS credentials with Bedrock access 3. Ensure IAM permissions for `bedrock:InvokeModel` --- ## 📦 Installation ### Quick Setup ```bash # Clone the repository git clone https://github.com/Tarique-B-DevOps/Agentic-Kubernetes-CLI.git cd Agentic-Kubernetes-CLI # Install dependencies pip install -r requirements.txt # Verify kubectl installation kubectl version --client # Run the agent python3 agent.py ``` --- ## ⚙️ Configuration ### Environment Variables Set or export the following variables before running the agent: ```bash # AWS Bedrock Configuration export BEDROCK_MODEL_REGION=us-east-1 export BEDROCK_MODEL_ID=apac.anthropic.claude-3-5-sonnet-20241022-v2:0 # AWS Credentials (optional if configured with AWS CLI) export AWS_ACCESS_KEY_ID=your_access_key export AWS_SECRET_ACCESS_KEY=your_secret_key export AWS_REGION=us-east-1 ``` --- ## 🎯 Usage ### Start the Agent ```bash python3 agent.py ``` ### Example Interactive Commands ```bash ⎈ kubectl> list all pods in production namespace ⎈ kubectl> what's wrong with my nginx deployment? ⎈ kubectl> show me the last 100 logs from api-server ⎈ kubectl> switch to staging cluster ``` --- ## 💡 Examples ### List Pods ```bash ⎈ kubectl> show all pods in the cluster ``` **AI Response:** ``` ✅ Command executed successfully: NAME READY STATUS RESTARTS AGE nginx-7854ff8877-2kxq9 1/1 Running 0 5d api-server-abc123 1/1 Running 2 3d ``` ### Troubleshoot a Pod ```bash ⎈ kubectl> why is my app-server pod failing? ``` **AI Response:** ``` Analyzing pod app-server... Issue: ImagePullBackOff Root Cause: Container exits with code 1 Last Error: Connection refused to database:5432 Suggestion: Check database service and credentials ``` ### View Logs ```bash ⎈ kubectl> get the last 50 lines of logs from nginx pod and summarize ``` ### Manage Contexts ```bash ⎈ kubectl> list all available clusters ⎈ kubectl> switch to production-cluster ⎈ kubectl> what's my current context? ``` ### Scale Deployments ```bash ⎈ kubectl> scale nginx deployment to 5 replicas ``` ### Namespace Operations ```bash ⎈ kubectl> show all resources in kube-system namespace ⎈ kubectl> get all deployments across all namespaces ``` --- ## 🏗️ Architecture ### System Components --- ## 🧩 MCP Tools ### run_kubectl_command ```python { "command": "kubectl get pods -n default -o json", "return_code": 0, "stdout": "...", "success": true } ``` ### kubectl_context ```python { "action": "list", "success": true, "output": "..." } ``` ## Source & license This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [Tarique-B-DevOps](https://github.com/Tarique-B-DevOps) - **Source:** [Tarique-B-DevOps/Agentic-Kubernetes-CLI](https://github.com/Tarique-B-DevOps/Agentic-Kubernetes-CLI) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/mcp-tarique-b-devops-agentic-kubernetes-cli - Seller: https://agentstack.voostack.com/s/tarique-b-devops - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.