k8s-mcp-server

Provides an MCP (Model Context Protocol) server exposing Kubernetes cluster interactions (discovery/listing/details/describe, logs, node/pod metrics, events, and resource create/update/delete), with optional Helm-related tooling and a read-only mode. Can run over stdio, SSE, or streamable-http transports.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ Infrastructure kubernetes mcp ai-agents observability helm devtools go cloud-infrastructure
⚙ Agent Friendliness
56
/ 100
Can an agent use this?
🔒 Security
47
/ 100
Is it safe for agents?
⚡ Reliability
31
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
85
Documentation
70
Error Messages
0
Auth Simplicity
75
Rate Limits
5

🔒 Security

TLS Enforcement
25
Auth Strength
55
Scope Granularity
40
Dep. Hygiene
40
Secret Handling
70

Runs as non-root in Docker containers per README. It supports Kubernetes auth via kubeconfig content, bearer token, or in-cluster service account, and offers --read-only plus tool category disabling to reduce risk. However, TLS/enforcement for the MCP HTTP/SSE endpoints is not clearly documented, and there is no described authentication/authorization protecting access to the MCP server itself; access control relies primarily on Kubernetes RBAC and how you expose the service.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
40
Error Recovery
30
AF Security Reliability

Best When

You have an MCP-capable agent and want a standardized, tool-based interface to Kubernetes (especially for listing/inspecting and read-only exploration, optionally with controlled write access).

Avoid When

When you cannot enforce strong Kubernetes RBAC, network isolation, and (if applicable) transport-level access controls for the MCP server endpoint.

Use Cases

  • Agent-assisted Kubernetes operations (read-only browsing, investigation, resource discovery)
  • Automating Kubernetes workflows such as creating/updating/deleting resources from YAML/JSON manifests
  • Observability via pod/node metrics and pod logs retrieved by an agent
  • Integrating Kubernetes tooling into an MCP-compatible assistant or web app

Not For

  • Highly sensitive production clusters without tight network/RBAC controls (it can modify resources)
  • Environments requiring strict, fine-grained authorization per tool-call beyond Kubernetes RBAC
  • Use as a public unauthenticated endpoint on the internet (no external auth described)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: KUBECONFIG_DATA (kubeconfig content via env) KUBERNETES_SERVER + KUBERNETES_TOKEN (bearer token) with optional CA/TLS settings In-cluster service account token from /var/run/secrets/kubernetes.io/serviceaccount/token Kubeconfig file via KUBECONFIG or default ~/.kube/config
OAuth: No Scopes: No

Authentication to Kubernetes is done via kubeconfig/token/service account. The README does not describe any separate auth for protecting access to the MCP server itself.

Pricing

Free tier: No
Requires CC: No

A hosted deployment is mentioned on a third-party site, but no pricing details are provided in the README excerpt.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Write-capable tools exist (resource create/update/delete and Helm install/upgrade/uninstall/rollback) unless mitigated via --read-only or tool disabling flags.
  • When using --no-k8s and --no-helm together, the server will exit with an error (must enable at least one tool category).
  • Authentication is to Kubernetes only; agents must ensure the MCP server endpoint is not exposed without appropriate access controls.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for k8s-mcp-server.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered