Security audit

xiaoliuren skills-小六壬占卜技能

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward Xiaoliuren divination knowledge skill with no hidden code execution, credential access, or privileged data handling.

Install only if you want a Chinese-language Xiaoliuren divination assistant. Treat any advice about health, money, legal matters, or major life choices as entertainment or cultural reference, not professional guidance.

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The README states that merely describing a divination request will cause the model to auto-load the skill, but it does not define strict trigger boundaries or exclusions. In agent ecosystems, vague activation criteria can cause unintended skill invocation during normal conversation, which widens the prompt surface and may override user expectations or route context into an unnecessary skill.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The Hermes section repeats broad auto-loading behavior based on a user's general description of a divination need, without defining scope limits or ambiguous-case handling. This can lead to unintended activation in loosely related chats, increasing the chance that the agent applies the skill where it was not requested or appropriate.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The guidance for other AI agents says they will automatically call the skill when users request Xiaoliuren divination, but it provides no implementation-specific guardrails. Because different agents interpret such instructions differently, this broad wording can encourage over-eager integrations that invoke the skill inappropriately or on weak semantic matches.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: The skill explicitly requires responses in Chinese without indicating that the assistant should adapt to the user's preferred language. This can cause usability and accessibility failures, and in multilingual deployments may lead to misunderstanding of important disclaimers or instructions. The content is not directly security-sensitive, but it violates safe skill design by constraining output language regardless of user context.

Natural-Language Policy Violations

Medium

Confidence: 98% confidence
Finding: The execution rules and output template hard-code Chinese headings and response structure, forcing Chinese-language output without user opt-in. This is risky because it can override the platform's normal language handling and reduce clarity for users who do not read Chinese, especially where the skill includes advice that should be clearly understood.

Static analysis

No suspicious patterns detected.