k8s-mcp-server

Provides a Model Context Protocol (MCP) server that exposes Kubernetes API functionality (read-only resource inspection plus optional mutation like deleting pods and scaling deployments) over MCP transports (stdio by default and SSE).

Evaluated Apr 04, 2026 (16d ago)
Repo ↗ Infrastructure ai-agents mcp kubernetes devops automation operations go
⚙ Agent Friendliness
58
/ 100
Can an agent use this?
🔒 Security
54
/ 100
Is it safe for agents?
⚡ Reliability
26
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
78
Documentation
70
Error Messages
0
Auth Simplicity
75
Rate Limits
10

🔒 Security

TLS Enforcement
45
Auth Strength
60
Scope Granularity
55
Dep. Hygiene
50
Secret Handling
60

Security depends primarily on Kubernetes RBAC least-privilege. README recommends restricted service accounts and read-only mode, which reduces risk. No explicit transport security details are provided for the SSE server (e.g., TLS requirements, auth for the MCP endpoint). No explicit error-handling/log redaction guidance or secret-handling guarantees are described in the provided content.

⚡ Reliability

Uptime/SLA
0
Version Stability
50
Breaking Changes
30
Error Recovery
25
AF Security Reliability

Best When

Used with least-privilege Kubernetes credentials and read-only mode, integrated locally via stdio or carefully secured via SSE in controlled environments.

Avoid When

Avoid running with broad RBAC or with --read-only=false unless you have strong operational controls and guardrails for agent-initiated actions.

Use Cases

  • Letting AI agents query Kubernetes cluster state (pods, deployments, services, configmaps, namespaces, nodes).
  • Debugging and troubleshooting workloads by combining cluster reads with agent workflows.
  • Automating operational tasks with controlled write access (e.g., scale deployments, delete pods).

Not For

  • Public/Internet-facing use without network controls and strict RBAC, since it can perform cluster mutations when read-only is disabled.
  • Handling sensitive cluster data without careful consideration of logging/telemetry and least-privilege credentials.
  • Replacing full operational tooling where robust audit/compliance workflows are required (not evidenced in the provided docs).

Interface

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

Authentication

Methods: Kubeconfig file (K8S_MCP_KUBECONFIG or --kubeconfig) In-cluster service account config (K8S_MCP_IN_CLUSTER or --in-cluster)
OAuth: No Scopes: No

Authentication/authorization is delegated to Kubernetes RBAC via the provided kubeconfig or in-cluster service account. No separate user/auth layer for the MCP server is described in the README.

Pricing

Free tier: No
Requires CC: No

Open-source MIT licensed; no pricing information in provided content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Default read-only is enabled ("default true"); agent requests that assume write access will fail unless read-only is disabled.
  • Cross-namespace and other wider permissions depend entirely on the kubeconfig/service account RBAC; mis-scoped credentials can broaden access.
  • Tool availability/coverage may depend on enabled resource types/toolsets via flags/env (K8S_MCP_RESOURCE_TYPES, K8S_MCP_TOOLSETS).

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 2026-04-04.

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