Security audit

Xue Feng Skill V192

Security checks for vulnerabilities and agentic risk

Overview

This career-planning skill is prompt-only and not technically malicious, but it requires sensitive personal details and uses gender-based steering for high-impact education and career advice.

Review before installing if you are uncomfortable with a skill asking for gender, family resources, connections, or political-screening-adjacent facts to shape education and career advice. Treat its recommendations as opinionated guidance, and avoid sharing sensitive personal details unless they are truly needed for your question.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
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)

Natural-Language Policy Violations

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill hard-codes sex-based recommendations such as different job-environment expectations for women and men, and specific steering like male nursing, without user opt-in or individualized necessity. That is dangerous because it operationalizes protected-attribute profiling and can systematically channel users toward or away from opportunities based on gender stereotypes rather than relevant qualifications or preferences.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: Making gender a mandatory intake field bakes protected-attribute personalization into every recommendation, even when it is not necessary to answer the user's question. In the context of this skill, that amplifies the risk from the later gendered rules because the system is instructed to always collect and use sex information as a decision input.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The activation rules are broad enough to trigger on general education or career discussions beyond narrow gaokao-major planning, which can cause the skill to intervene unexpectedly and apply its prescriptive framework in contexts the user did not intend. In this skill, that matters because the framework includes sensitive intake and strong normative recommendations, so over-activation increases the chance of unnecessary profiling or low-quality guidance.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill requires collecting sensitive or high-risk personal attributes such as gender, family conditions, industry connections, desired city, and even elsewhere discusses criminal-record/political-screening issues, but it does not tell users these fields are optional, why they are needed, or how to proceed without them. This creates privacy and fairness risk because users may feel compelled to disclose more than necessary for advice.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The skill name and description are entirely in Chinese, and the metadata does not indicate that users can choose another language or that the skill is limited to a Chinese-speaking audience. Under the language/locale policy, a forced language without user opt-in can be a natural-language policy violation.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.