huawei-cloud-lts-log-inspector

• Outcome: Review
• Latest audit: Sep 3, 2026 · 12:26 PM UTC
• Version: 1.0.0

Overview

The skill is coherent for Huawei Cloud LTS log inspection, but its setup and credential guidance are too weak for a tool that can read/export logs and manage transfer tasks.

Review the CLI installer source and verify it through a trusted Huawei channel before running it. Use least-privilege, preferably short-lived credentials, avoid placing AK/SK secrets in persistent shell profiles, and require explicit confirmation before any OBS transfer create or delete operation.

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection

Findings (10)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 95% confidence
• Finding: The skill documents multiple shell-based `hcloud` commands but does not declare any `permissions` or `allowed-tools` scope. In an agent environment, this can let the runtime expose shell execution more broadly than intended, increasing the chance of command execution against real cloud resources without explicit least-privilege boundaries.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The manifest-style description lists triggers such as "故障深度定位", "log context", and "collection inspection" without restricting them to Huawei Cloud LTS context. These phrases are generic enough to overlap with ordinary debugging or inspection requests, which could cause unintended invocation.

Intent-Code Divergence

Medium

• Confidence: 92% confidence
• Finding: The skill says bulk export of all logs is prohibited, but the operational examples do not encode enforceable safeguards beyond a narrative statement. A user or downstream agent could still invoke transfer creation with broad or unintended scope, leading to excessive data export, privacy exposure, and avoidable storage/transfer cost.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The guide instructs users to fetch and immediately execute a remote shell script via curl and bash without any integrity verification, signature check, or out-of-band trust validation. This creates a software supply chain risk: if the remote object, hosting bucket, DNS path, or TLS trust chain is compromised, users could execute arbitrary code on their systems.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The guide recommends exporting long-lived cloud access credentials into environment variables without warning about secret handling risks. Environment variables can be exposed through shell history, process inspection, debugging output, CI logs, crash reports, or inherited subprocess environments, which can lead to credential theft and unauthorized cloud access.

Autonomous Decision Making

Medium

• Category: Excessive Agency

Content

### Query Commands (Read-Only)

Query commands are safe to execute without confirmation. Verify results by
checking the response structure:

```bash

• Confidence: 75% confidence
• Finding: Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Natural-Language Policy Violations

Medium

• Confidence: 88% confidence
• Finding: This manifest/config file contains locale- and region-specific defaults such as "cn-north-4" and "Asia/Shanghai" that constrain execution to a specific geography. Under the policy, forcing a specific locale or region without user opt-in or an explicit documented justification is a natural-language policy concern.

Natural-Language Policy Violations

Low

• Confidence: 86% confidence
• Finding: The command examples fix `--timezone=Asia/Shanghai`, which imposes a specific locale assumption in natural-language guidance. The file does not offer the user a timezone choice or explain that the skill is intentionally region-specific beyond a generic region example.

Natural-Language Policy Violations

Low

• Confidence: 84% confidence
• Finding: This diagnostic workflow repeats `--timezone=Asia/Shanghai` as a fixed value, reinforcing a locale-specific default. Because the skill otherwise presents itself as a general Huawei Cloud LTS tool, this can conflict with language/locale choice expectations.

Natural-Language Policy Violations

Low

• Confidence: 84% confidence
• Finding: The full patrol report example again uses `--timezone=Asia/Shanghai` with no opt-in or explanation. Repeated fixed locale guidance across the skill increases the likelihood that outputs will be anchored to an unintended region.

Static analysis

No suspicious patterns detected.