1. Conduid
  2. AI
  3. Kdebug MCP
MCP server · AI

Kdebug MCP

KDebug is a Kubernetes debugging tool that allows you to interact with your Kubernetes clusters through LLMs. It uses the Model Control Protocol (MCP) to enable AI to execute Kubernetes commands on your behalf.

Unclaimed last commit a year ago ai
46Fair

Scored 4 days ago · breakdown

About Kdebug MCP

Kdebug MCP is an MCP server published by soub4i in the AI category: kDebug is a Kubernetes debugging tool that allows you to interact with your Kubernetes clusters through LLMs. It uses the Model Control Protocol (MCP) to enable AI to execute Kubernetes commands on your behalf.

The repository has 16 stars and 1 forks, with the last commit a year ago. Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.

Install

From source
git clone https://github.com/soub4i/kdebug-mcp

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about Kdebug MCP

Powered by Claude · Grounded in docs

I know everything about Kdebug MCP. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

Static checks derived from the repository 5 days ago. Heuristics on the source, not an audit.

  • ✓README presentThe repository ships documentation.
  • !License declaredNo license file. Legally you have no right to use it in a product.
  • !Tests presentNo tests found. Behaviour changes may go unnoticed between versions.
  • ✓Dependencies pinnedA lockfile pins dependency versions.
  • ✓No dynamic code executionSource scan found no eval, dynamic Function, or shell execution.
  • !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.

Releases

v0.0.1Release v0.0.1 · 27 Apr 2025First one

README

KDebug - Kubernetes Debugging MCP Server

KDebug is a Kubernetes debugging tool that allows you to interact with your Kubernetes clusters through Claude AI. It uses the Model Control Protocol (MCP) to enable Claude to execute Kubernetes commands on your behalf.

Overview

This tool allows you to:

  • Inspect Kubernetes resources (pods, services, deployments, etc.)
  • View pod logs
  • Check node status
  • Monitor events
  • And more, all through conversational AI

Installation

Prerequisites

  • Go 1.18+ (if building from source)
  • Kubernetes cluster access configured (~/.kube/config)
  • Claude client with MCP support

Option 1: Download pre-built binary

  1. Go to the GitHub releases page:
https://github.com/soub4i/kdebug-mcp/releases
  1. Download the latest release for your platform (macOS, Linux, Windows)

  2. Make the binary executable:

chmod +x kdebug-mcp
  1. Move the binary to a location in your PATH:
mv kdebug-mcp /usr/local/bin/kdebug-mcp
# or 
mv kdebug-mcp ~/bin/kdebug-mcp  # If you have ~/bin in your PATH

Option 2: Build from source

  1. Clone the repository:
git clone https://github.com/soub4i/kdebug-mcp.git
cd kdebug-mcp
  1. Build the binary:
go build -o bin/server ./cmd/server/main.go

Configuration

Configure Claude to use KDebug

Create or edit the Claude MCP configuration file located at:

  • macOS: ~/Library/Application Support/com.anthropic.claude/config.json
  • Linux: ~/.config/com.anthropic.claude/config.json
  • Windows: %APPDATA%\com.anthropic.claude\config.json

Add the following configuration:

{
    "mcpServers": {
        "kdebug": {
            "command": "/path/to/kdebug-mcp/bin/server"
        }
    }
}

Replace /path/to/kdebug-mcp/bin/server with the actual path to your KDebug binary.

Kubernetes Context

KDebug uses your current Kubernetes context from ~/.kube/config. Make sure your Kubernetes configuration is properly set up.

To switch contexts, you can use:

kubectl config use-context <context-name>

Usage

  1. Start Claude and make sure it's connected to your KDebug MCP server
  2. In your conversation with Claude, ask about your Kubernetes resources

Example prompts:

  • "what's wrong with my cluster.context: minikube. namespace: default."
  • "Show me all pods in the default namespace"
  • "What services are running in the kube-system namespace?"
  • "Get the logs from pod xyz in namespace abc"
  • "List all nodes in my cluster"
  • "Check for recent events in the default namespace"

Available Commands

KDebug provides access to the following Kubernetes resources:

  • nodes: List all nodes in the cluster
  • pods: List pods in a namespace or get a specific pod
  • podLogs: Get logs from a specific pod
  • services: List services in a namespace or get a specific service
  • deployments: List deployments in a namespace or get a specific deployment
  • statefulsets: List stateful sets in a namespace or get a specific stateful set
  • replicasets: List replica sets in a namespace or get a specific replica set
  • daemonsets: List daemon sets in a namespace or get a specific daemon set
  • events: List events in a namespace or related to a specific resource

Troubleshooting

Common Issues

  1. Claude can't connect to KDebug

    • Check that the path in your Claude MCP configuration is correct
    • Ensure the KDebug binary is executable
  2. Permission errors

    • Make sure your Kubernetes configuration (~/.kube/config) has the necessary permissions
    • Try running kubectl get pods to verify your Kubernetes access
  3. Context switching

    • If you have multiple Kubernetes contexts, Claude will ask which context to use
    • Ensure you have access to the context you're trying to use

Security Considerations

KDebug executes Kubernetes commands with the permissions of your current user. Be mindful that Claude will have the same access to your Kubernetes cluster as you do through your configured kubectl context.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

README fetched from the repository 5 days ago. The original is authoritative.

Questions

About Kdebug MCP

How do I install Kdebug MCP?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is Kdebug MCP safe to use with an AI agent?

Its trust score is 46 out of 100 (fair). It passes 3 of the 6 static security checks run so far; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Kdebug MCP still maintained?

The last commit was a year ago. That's long enough that you should check whether the maintainer is responding to issues before depending on it.