mirrord-operator
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- COMMAND_EXECUTION (HIGH): The skill instructs the agent to perform high-privilege operations on a Kubernetes cluster, including namespace creation (
kubectl create namespace), operator installation (helm install), and RBAC modification (ClusterRolecreation). Improper handling of these commands poses a significant risk to cluster integrity. - PROMPT_INJECTION (HIGH): Indirect Prompt Injection Surface (Category 8).
- Ingestion points: The skill processes untrusted user input for license keys, target pod names, and Helm configuration values.
- Boundary markers: Absent. There are no delimiters or instructions to ignore malicious content embedded within user-provided strings.
- Capability inventory: The skill possesses extensive capabilities via
kubectlandhelmto modify cluster state, install packages, and view logs. - Sanitization: Absent. There is no evidence of validation or escaping for user-supplied arguments before they are passed to shell commands.
- EXTERNAL_DOWNLOADS (MEDIUM): The skill downloads and installs Helm charts from an external repository (
https://metalbear-co.github.io/charts). This source is not included in the pre-defined trusted organizations list, making the dependency unverifiable for safety.
Recommendations
- AI detected serious security threats
Audit Metadata