How to Configure Continue Kubernetes Audit Log
To set up a continue kubernetes audit log workflow, you bind a scoped role to the coding tool and turn on API server logging. This records every cluster request the coding assistant makes on behalf of a developer. Engineering teams often connect local coding assistants to test or live clusters. Developers want the assistant to read pod status, check events, or run commands. To make this work fast, teams often give the tool a wide cluster role. When roles are too broad, a single bad prompt can c
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 a continue kubernetes audit log workflow, you bind a scoped role to the coding tool and turn on API server logging. This records every cluster request the coding assistant makes on behalf of a developer.
Engineering teams often connect local coding assistants to test or live clusters. Developers want the assistant to read pod status, check events, or run commands. To make this work fast, teams often give the tool a wide cluster role. When roles are too broad, a single bad prompt can change live workloads or read sensitive secrets.
How the native continue kubernetes audit log setup works
Kubernetes controls API requests through Role-Based Access Control, known as RBAC. In this model, a Role or ClusterRole holds rules that list allowed actions. Permissions in Kubernetes are only additive, which means there are no deny rules. A Role sets permissions inside one namespace. A ClusterRole covers the whole cluster, including nodes or resources across all namespaces.
To let the Continue coding tool talk to the cluster, you create an identity for it. You then link that identity to a Role using a RoleBinding in one namespace. If the tool needs cluster-wide reach, you use a ClusterRoleBinding instead. Every time the assistant queries the cluster API, the API server checks these bindings.
Some requests target subresources like pod logs or pod command execution. In RBAC rules, you write these subresources with a slash after the main resource name. The API server logs each request and response in its audit log. Security teams then read this log to see what the coding assistant did.
Where native roles and audit logs fall short
Native RBAC and audit logs solve the basic problem of who can call the API. They let you limit a coding assistant to one namespace and record the API call history. Still, this setup only checks permissions when the API call starts. It does not inspect what happens inside an active command session or what data returns.
For example, if you grant pod execution rights, the assistant can run shell commands inside a container. The native audit log shows that an execution call started, not the full output stream. Similarly, if the assistant reads pod logs, those logs might hold customer emails, phone numbers, or card numbers. Native RBAC cannot mask those fields in flight before they reach the model.
Audit Log Integrity + Kubernetes Audit Logs: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
Developers also store long-lived cluster keys on their laptops so the local assistant can work. If a laptop is lost or a prompt goes wrong, those local keys stay valid until someone revokes them. Teams need a way to inspect commands in flight, remove local keys, and mask sensitive fields like national IDs in results.
Scoping permissions for coding assistants
When you set up native roles for a coding assistant, start with the smallest possible scope. Use a namespace Role and RoleBinding instead of cluster-wide rules whenever you can. Only grant read actions on pods and events if the assistant just helps debug apps.
Be very careful with subresources that allow remote command execution or log reading. Because RBAC rules are additive, adding a wide role later will override your tight restrictions. Review your bindings often to make sure test permissions did not leak into live namespaces. Pair these role checks with clear audit rules so your team knows which user ran each request.
Check Kubernetes API requests and responses on the wire
Kubernetes RBAC decides who may call the API server. RBAC does not check response payloads or split listing Secret names from reading Secret values when a role grants get access. With hoop.dev, you run an open-source HTTP proxy, a sidecar for your Kubernetes API. The kubectl client changes only the server address it talks to while RBAC stays in place.
On each call, hoop.dev guardrails read the method, normalized path, and headers to refuse bad deletes or full Secret reads with a 403 error. On the way back, hoop.dev data masking walks JSON responses by key and masks matched values in Secrets, ConfigMaps, and pod logs. Every request and verdict is logged without storing bearer tokens. View the project on GitHub.
Frequently asked questions
How does Kubernetes handle deny rules for coding assistants?
Kubernetes RBAC does not have deny rules because all permissions are additive. You must start with zero permissions and only grant the exact actions the assistant needs.
Should I use a RoleBinding or a ClusterRoleBinding for Continue?
Use a RoleBinding whenever possible so the coding assistant stays inside one single namespace. A ClusterRoleBinding gives access across the whole cluster, which increases risk if a prompt runs a bad command.
Can native Kubernetes audit logs hide sensitive data from pod logs?
No, native audit logs only record API requests and do not rewrite pod log content. If pod logs contain emails or card numbers, native tools will not mask them in the results.
Related guides
In this post