advanced-kubernetes
Warn
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The
scripts/setup-operator-dev.shscript executessudo mvto install thekubebuilderbinary into a system path (/usr/local/bin/). While common for CLI tools, running scripts that require root privileges requires careful review. - [UNVERIFIABLE_DEPENDENCIES]: The skill downloads and installs tools from external domains during the environment setup process.
- Evidence:
curl -L -o kubebuilder https://go.kubebuilder.io/dl/latest/...andkubectl apply -f https://github.com/cert-manager/cert-manager/releases/download/...inscripts/setup-operator-dev.sh. - Context: The download sources (Kubebuilder and Cert-Manager) are well-known projects within the Kubernetes ecosystem, which mitigates the risk according to trusted source rules.
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for building controllers that ingest data from the Kubernetes API server (Custom Resources). This creates a surface for Indirect Prompt Injection if an agent later processes this data.
- Ingestion points:
controllers/myapp_controller.go(reading MyApp resources viar.Get). - Boundary markers: No explicit prompt delimiters for the data in the logic, but the data is structured (OpenAPI v3).
- Capability inventory: The provided controller templates can create and manage Deployments and Services (
controllers/myapp_controller.go). - Sanitization: The templates include OpenAPI validation schemas in
templates/crd-definition.yamland admission webhook logic intemplates/webhook.goto validate inputs.
Audit Metadata