Huawei Cloud Swr Image Governance

• Outcome: Review
• Latest audit: Jun 17, 2026 · 2:33 AM UTC
• Version: 0.1.0

Overview

The skill is coherent, but it can make persistent Huawei Cloud permission, sharing, delegation, and cleanup changes without enough guardrails for production use.

Install only if you intend to let an agent help administer Huawei Cloud SWR. Use a dedicated least-privilege IAM user, review every write command before execution, start with read-only audits, and require explicit confirmation before creating agency delegation, granting manage/edit access, creating shared domains, or changing retention rules that could delete tags.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (7)

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The guide documents `CreateAgency` with only a brief use-case note and no prominent warning that it grants SWR delegated access to other cloud services on the user's behalf. In a governance skill focused on permissions and repositories, omitting a clear consent and scope warning can mislead users into enabling cross-service trust they do not fully understand, increasing the risk of privilege expansion or unintended access paths.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill documents `CreateAgency` as a routine setup step and even recommends invoking it when `is_agency` is false, but it does not clearly warn that this changes IAM-style delegation and grants SWR additional access to other services such as OBS or CCE. In an agent setting, this can normalize a permission-impacting action as a harmless check/fix flow, increasing the chance of unintended privilege expansion or unauthorized cross-service access if run without explicit user confirmation.

Missing User Warnings

Medium

• Confidence: 87% confidence
• Finding: The file provides direct grant, modify, and revoke commands for repository access control, including manage-level permissions, but does not explicitly warn that these actions can expose private images, enable unauthorized pushes, or disrupt legitimate access. In a permission-management skill, omission of security impact guidance increases the chance of accidental over-privileging or improper revocation by users following examples verbatim.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The document states that tags not matching any rule condition will be deleted, but the examples do not prominently warn that a misconfigured selector or overly narrow pattern can cause destructive cleanup of needed image tags. In an image-governance skill, this is dangerous because users may copy commands directly into production repositories and unintentionally trigger data-loss-like deletion behavior during automated retention execution.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: Updating a retention rule changes future deletion criteria, but the workflow does not warn that tags previously retained may become eligible for cleanup after the update. This creates a realistic risk of accidental destructive changes, especially when reducing retention counts or narrowing selectors in repositories that users assume are already safe.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill provides step-by-step instructions to create shared download domains but does not clearly warn that this grants external parties a path to pull repository images without normal repository permissions. In an image-governance skill focused on access control, omission of this security context can lead users to unintentionally expose sensitive or internal container images beyond the intended trust boundary.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The verification guide instructs users to create and delete namespace permissions against a potentially real SWR namespace without any explicit warning that these steps modify live access controls. In a verification context, operators may run the commands in production or against shared environments, unintentionally granting repository access to another IAM user or briefly altering authorization state.

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.