huawei-cloud-ucs-cluster-onboarding-manager

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to download the Huawei Cloud KooCLI (hcloud) installation script from an official vendor domain (obs.cn-north-4.myhuaweicloud.com). This is a standard setup procedure for the vendor's toolset.
  • [REMOTE_CODE_EXECUTION]: The CLI installation guide includes a pattern to download and execute a shell script using curl and bash. Since the source is a verified vendor repository (huaweicloud), this is considered a safe and intended operational step.
  • [COMMAND_EXECUTION]: The skill uses the hcloud command-line interface to interact with cloud APIs. It includes an automated acceptance of privacy terms using printf "y\n" | hcloud version, which is a common automation pattern for CLI tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data including Kubernetes cluster metadata, annotations, and kubeconfig content when registering clusters.
  • Ingestion points: Cluster registration parameters in task-cluster-registration.md and SKILL.md specifically metadata.name and metadata.annotations.kubeconfig.
  • Boundary markers: None identified for individual parameter inputs.
  • Capability inventory: The skill can create, update, and delete cloud resources and retrieve sensitive kubeconfig files via the hcloud CLI.
  • Sanitization: The skill relies on the underlying cloud provider's API and CLI validation to handle input data. Potential risks are mitigated by the requirement for explicit user confirmation before executing billing-related or destructive actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 08:20 PM
Security Audit — agent-trust-hub — huawei-cloud-ucs-cluster-onboarding-manager