How to Configure Make Kubernetes Permissions
To set up make kubernetes permissions, you create a native role with specific rules. You then bind that role to the account used by your Make workflow. Many teams start by giving their workflow a wide admin role to save time. This broad setup creates big risks if a bad task runs in your cluster. If an attacker steals that key, they can change or delete any app. How make kubernetes permissions work in native RBAC Kubernetes uses a built-in role system to control who can do what. Every rule in
Your coding agents run with your access.
Claude Code reaches prod with a developer's credentials, and approval prompts get clicked through. hoop.dev checks every query, masks sensitive data, and holds risky changes for a human.
To set up make kubernetes permissions, you create a native role with specific rules. You then bind that role to the account used by your Make workflow. Many teams start by giving their workflow a wide admin role to save time. This broad setup creates big risks if a bad task runs in your cluster. If an attacker steals that key, they can change or delete any app.
How make kubernetes permissions work in native RBAC
Kubernetes uses a built-in role system to control who can do what. Every rule in this system is purely additive. There are no deny rules in native Kubernetes role controls. You start with zero access and add rules for each needed action.
First, you pick between a namespace role and a cluster-wide role. A standard role sets rules inside one single namespace. A cluster role covers the whole cluster and all namespaces at once. It can also cover cluster-scoped items like nodes.
For a Make workflow, you should almost always pick a single namespace role. You list the exact items the workflow needs to read or change. You also need to list subresources if the workflow reads pod logs. In Kubernetes, pod logs and pod exec are subresources under the main pod item.
Next, you link the role to an account with a binding. A role binding grants those rules inside one namespace. It can link to a standard role or a shared cluster role. A cluster role binding gives access across the entire cluster at once. Once you save the binding, the API server checks every Make request against it.
What native roles solve and where gaps remain
Native roles and bindings solve the basic problem of who can log in. They stop a Make workflow from touching namespaces it does not own. They also block the workflow from deleting nodes or reading cluster secrets. If a workflow only needs to read pods, native rules enforce that limit well.
Still, native rules only check the resource type and the action verb. They do not inspect what happens inside an allowed request in flight. For example, a workflow might need the pod exec subresource to run a check. Native rules allow or block pod exec as a whole.
Kubernetes RBAC + AI Agent Permissions: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Once the API server allows the exec call, Kubernetes stops checking the commands. The workflow can run any shell command inside that container. Native roles also cannot mask sensitive data returned in pod logs or outputs. If a log line holds emails, phone numbers, or card numbers, Make receives them raw. Finally, static account keys stored in external tools are hard to rotate often.
Best practices for scoping workflow roles
Start every new Make workflow with a fresh account in one namespace. Never reuse a human admin key or a shared default account. Write down the exact API calls your workflow makes during a normal run.
Put only those specific items and verbs into a single namespace role. Avoid wildcards in your rules, as they grant access to future items too. Keep subresources like pod exec out of the role unless required. Review your bindings each month and remove any rules the team no longer uses. If a workflow needs to read sensitive output, add controls that rewrite matched fields in results.
Add request guardrails and response masking beside Kubernetes
Cluster roles and service accounts govern who reaches the Kubernetes API. Still, a broad role or agent can delete workloads or print sensitive values from ConfigMaps and logs. In practice, hoop.dev runs beside the API server as a sidecar for your Kubernetes API and checks every request and response alongside RBAC.
With hoop.dev Kubernetes controls, guardrails can allow listing Secret names while refusing reads of Secret data before execution. Data masking redacts matched keys in returned JSON responses and log streams in memory. Furthermore, hoop.dev holds SOC 2 Type II and is MIT licensed on GitHub.
Frequently asked questions
Can I write a deny rule to block one specific pod?
No, Kubernetes permissions are purely additive and do not support deny rules. You must grant access only to the exact items the workflow needs.
Should Make use a RoleBinding or a ClusterRoleBinding?
Use a RoleBinding whenever your workflow operates inside a single namespace. Reserve ClusterRoleBinding for rare tasks that must read or change cluster-wide items.
How do I let a Make workflow read pod logs safely?
Grant get and list rules on the pods log subresource inside a namespace role. Do not grant full pod write or exec rights just to read logs.
Related guides
In this post