// SCAN REPORT

K8s

41 tools. 9 can modify or destroy data without limits.

2 destructive tools with no built-in limits. Policy required.

Last updated:

9 can modify or destroy data
32 read-only
41 tools total

Community server · catalogue entry verified 29/06/2026

How to control K8s ↓

TOOL MAP

What K8s exposes to your agents

Read (32) Write / Execute (7) Destructive / Financial (2)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS
Critical Risk

The most dangerous K8s tools

Destructive · High delete_pod Deleting a pod terminates the running container(s) and workload associated with it. Destructive · Critical delete_resource Deleting Kubernetes resources (pods, deployments, services, etc.) is irreversible and has immediate blast radius effects: workloads stop running, data may be lost, services become unavailable, and clusters destabilize. Execute · High restart_deployment restart_deployment executes a concrete action against a Kubernetes cluster—initiating a rolling update of a deployment. Execute · High scale_deployment Scaling a deployment is an Execute action because it triggers external operations (modifying the Kubernetes cluster state) whose blast radius depends on the replica count argument—a misconfigured scale-up could exhaust r

9 of K8s's 41 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control K8s

PolicyLayer is an MCP gateway — it sits between your AI agents and K8s, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations
{
  "delete_pod": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations
{
  "apply_manifest": {
    "limits": [
      {
        "counter": "apply_manifest_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations
{
  "describe_resource": {
    "limits": [
      {
        "counter": "describe_resource_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

  1. Create a free account and register K8s — nothing to install.
  2. Add these rules — paste them, or build them visually. Tune the limits to your setup.
  3. Point your MCP client (Claude, Cursor, anything) at your gateway URL.
ENFORCE POLICY ON K8S →

Instant setup, no code required.

FULL CATALOGUE

All 41 K8s tools

DESTRUCTIVE 2 tools
Destructive delete_pod Delete a pod. Destructive delete_resource Delete a Kubernetes resource by type and name.
EXECUTE 5 tools
Execute apply_kustomize Render and apply a Kustomize directory. Execute exec_command exec_command Execute restart_deployment Restart a deployment via rolling update (equivalent to kubectl rollout restart). Execute scale_deployment Scale a deployment to the specified replica count. Execute wait_for_ready wait_for_ready
WRITE 2 tools
Write apply_manifest Apply a Kubernetes YAML manifest (create or update resources). Write generate_deploy_manifests generate_deploy_manifests
READ 32 tools
Read describe_resource describe_resource Read diagnose_pod diagnose_pod Read get_configmap Get a ConfigMap's metadata and data contents. Read get_contexts List available kubeconfig contexts and indicate the active one. Read get_current_context Get the currently active kubeconfig context with cluster and user details. Read get_deployment Get detailed information about a deployment. Read get_ingress Get detailed information about an Ingress, including rules and TLS config. Read get_node Get detailed information about a node. Read get_pod Get detailed information about a specific pod. Read get_pod_logs get_pod_logs Read get_resource_yaml get_resource_yaml Read get_role Get detailed information about a Role or ClusterRole. Read get_role_binding Get detailed information about a RoleBinding or ClusterRoleBinding. Read get_rollout_status Check the rollout status of a deployment. Read get_secret get_secret Read get_service Get detailed information about a service. Read get_service_account Get detailed information about a ServiceAccount. Read list_configmaps List ConfigMaps in a namespace or across all namespaces. Read list_deployments List deployments in a namespace or across all namespaces. Read list_events List cluster events, optionally filtered by involved resource. Read list_ingresses List Ingresses in a namespace or across all namespaces. Read list_jobs List jobs in a namespace or across all namespaces. Read list_namespaces List all namespaces in the cluster. Read list_nodes List all cluster nodes. Read list_pods List pods in a namespace or across all namespaces. Read list_role_bindings list_role_bindings Read list_roles list_roles Read list_secrets List Secrets in a namespace or across all namespaces. Read list_service_accounts List ServiceAccounts in a namespace or across all namespaces. Read list_services List services in a namespace or across all namespaces. Read top_nodes Show CPU and memory usage for nodes (requires metrics-server). Read top_pods top_pods
EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

Yaver 1060 tools
adminautomation
Nodebench 824 tools
adminautomation
UnClick 813 tools
adminautomation
Binance MCP Server 734 tools
adminautomation
Fortimanager 584 tools
adminautomation
Crow 573 tools
adminautomation
Mcp 571 tools
adminautomation
GoHighLevel MCP Server 566 tools
adminautomation
FAQ

Questions about K8s

Can an AI agent delete data through the K8s MCP server? +

Yes. The K8s server exposes 2 destructive tools including delete_pod, delete_resource. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through K8s? +

The K8s server has 2 write tools including apply_manifest, generate_deploy_manifests. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach K8s.

How many tools does the K8s MCP server expose? +

41 tools across 4 categories: Destructive, Execute, Read, Write. 32 are read-only. 9 can modify, create, or delete data.

How do I enforce a policy on K8s? +

Register the K8s MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every K8s tool call.

Deterministic rules across all 41 K8s tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON K8S →

Instant setup, no code required.

41 K8s tools catalogued and risk-classified — across an index of 43,000+ MCP servers.

// GET IN TOUCH

Have a question or want to learn more? Send us a message.

Message sent.

We'll get back to you soon.