Security audit

Qian Jiu Jiu

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Chinese typo and grammar preprocessor with no executable code, network access, credential use, or hidden persistence.

Install this only if you want every message in relevant Chinese-language conversations checked for obvious wording issues before the agent proceeds. Expect occasional confirmation prompts, but the artifact does not show data exfiltration, credential access, destructive behavior, or hidden execution.

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)

Natural-Language Policy Violations

Medium

Confidence: 91% confidence
Finding: The README states the skill requires 'Simplified Chinese language support' and all examples/instructions assume Chinese-only operation. This creates a language-policy concern because the skill appears to force a specific language context without documenting user choice, opt-in, or a clearly justified region-specific limitation.

Natural-Language Policy Violations

Medium

Confidence: 85% confidence
Finding: The documentation indicates Chinese-language handling as the default operating mode, but it does not describe a clear user opt-in before analyzing message content. This makes the skill more likely to process user input unexpectedly, especially in mixed-language conversations, and may introduce privacy, usability, and consent issues when users did not intend preprocessing.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The skill states it automatically activates on each user message, which creates an overly broad interception scope for all conversations rather than a clearly scoped or user-invoked workflow. In practice this can cause unintended preprocessing of unrelated content, alter conversation flow through confirmation gating, and expose users to silent behavior changes without explicit consent.