Audit advanced Works with: Any AI

Kubernetes Manifest Review

Table output Has example

Check resource limits, probes and privileges.

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

Review Kubernetes manifests. Paste deployments, services and ingress definitions with secrets removed: [MANIFESTS]. Cluster setup and who has access: [CLUSTER_ACCESS]. Incidents seen: [INCIDENTS].

Check resource requests and limits, probes, security contexts, privilege settings, network policies, service accounts, secrets handling, image tags, update strategy and disruption budgets. Explain each in operational terms.

Return a table with Resource, Issue, Severity (red, amber, green) and Fix, a corrected manifest for the riskiest deployment, and a staged rollout plan.

Also suggest a baseline policy for new workloads, a way to check manifests in continuous integration, and a list of settings that should never be set without review. Add a plan for testing changes in a non-production cluster before they reach production.

Finish with the three settings to fix before the next release.

Then turn the findings into a platform baseline and rollout. Write a policy set in plain language covering resource limits, security contexts, network policies, image sources, labels, probes and disruption budgets, each with the reason, an exception process and a way to test compliance in continuous integration. Provide a staged adoption plan: warn only for two weeks, then enforce for new workloads, then fix existing ones in order of risk. Create a sample hardened manifest for a typical stateless service with comments, and a checklist for stateful services that need extra care. Add a section on secrets management and service account permissions, and on how to roll back a failed change. List the metrics and events to watch after tightening, such as restarts and throttling. Keep the guidance defensive and test in a non-production cluster before changing production.

Do not claim the system is secure. State what was and was not reviewed, list assumptions, and remind me that this review does not replace a penetration test or compliance audit by qualified people.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

Resource table (invented deployment)
Container resources: none set. Amber (risk of noisy neighbours).
Probes: liveness only; readiness missing, so traffic arrives before startup. Red.
Security context: runs as root, writable root file system. Red.
Image tag: latest. Red.
Disruption budget: none. Amber.
Network policy: none; all pods can reach the database. Amber.
Corrected manifest outline: requests 250m CPU and 256Mi memory; limits as agreed; readiness at /ready; runAsNonRoot true; read-only root file system; pinned image digest.
Rollout: apply in a non-production cluster first; watch restarts and throttling for a day.

• audit