Security audit

Xungen

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese-language entertainment and cultural-history skill that generates playful Shang-dynasty ancestry reports from a surname or hometown, with no code execution or hidden access.

Install this if you want a Chinese-language, playful Shang-dynasty ancestry-style report. Treat outputs as cultural entertainment rather than verified genealogy, and use the explicit /xungen command if you want to avoid accidental activation from ordinary surname or hometown mentions.

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 (3)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The trigger conditions are broad enough to activate on ordinary conversation about surnames, ancestry, or hometowns, which can cause unintended skill invocation and scope hijacking. In an agent setting, that can override the user’s actual intent, inject long unsolicited outputs, or cause the system to process sensitive personal/family context when the user did not explicitly request this skill.

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The usage guidance includes vague phrases like directly saying '我姓X' or asking to check ancestors, which are common in normal dialogue and can lead to accidental activation. This increases the risk of prompt-routing errors, where the agent enters this skill instead of answering the user’s broader question or handling the information more minimally.

Natural-Language Policy Violations

Low

Confidence: 82% confidence
Finding: The manifest and full instructions are entirely in Chinese and specify trigger phrases, outputs, and interaction patterns only in Chinese, with no indication that users may choose another language. For a general-purpose skill, forcing one language without opt-in can violate language/locale policy expectations.

Static analysis

No suspicious patterns detected.