Security audit

超能去AI

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese text rewriting skill with disclosed creative augmentation risks, but no hidden code execution, network access, credential access, or persistence beyond optional output text.

Install only if you want a Chinese creative or informal rewriting assistant. Do not use it for academic work, legal text, news, technical documentation, or any text where exact facts, numbers, attribution, or authorship transparency matter, because the skill may add or reshape details to make the writing feel human.

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

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (19)

Description-Behavior Mismatch

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill presents itself as a rewrite tool but explicitly instructs the model to add invented narrative elements such as '无用道具', '干扰项', human-grayness, and personal memory anchors. This creates a semantic integrity risk: users may believe the output remains a faithful paraphrase when the skill is actually fabricating details, which is especially dangerous if the text is later reused outside purely creative contexts.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The file claims '不改事实' and '不改核心意思', yet nearby hard constraints require transformations that can alter factual precision and introduce unsupported content, such as converting precise numbers to body-feel descriptions and adding unrevealed distractors or new personal details. These conflicting instructions can cause the model to produce misleading outputs while falsely signaling fidelity, undermining trust and enabling accidental misinformation.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger list includes common conversational phrases such as '帮我改一下，太像AI写的了' and '这看着太AI了' that could easily appear in ordinary user requests. In an agent environment, overly broad activation phrases can cause unintended invocation of the skill, leading to unauthorized content transformation, loss of user intent fidelity, or accidental application in contexts where rewriting to conceal AI authorship is inappropriate.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The description states "专为中文文本设计," which imposes a language constraint in the skill's natural-language behavior. The file does not offer a user choice, opt-in, or a documented policy justification for restricting operation to Chinese, so it creates a locale/language policy concern under the stated rule.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 86% confidence
Finding: The trigger phrases are broad colloquial expressions like '帮我改一下，太像AI写的了' and '太像AI了帮我改改', which can appear in normal conversation and cause the skill to activate unintentionally. Because this skill can substantially alter text structure and inject new content, accidental invocation increases the chance of undesired or silent modification.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The skill routes technical, business, workplace, and science-like content into rewrite roles even though those domains are fact-sensitive and vulnerable to subtle distortion. Even with the '语言层改写' limitation, the prescribed transformations can still change nuance, certainty, terminology, or precision in ways that materially affect meaning.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The instruction says that when the source tone is unclear, the editor should default to plain spoken Chinese. This is a natural-language locale policy constraint, and the file does not provide user opt-in, alternative language handling, or a documented regional justification.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The entire role definition is written as a prescriptive persona in Chinese and provides no indication that language choice is optional or user-selectable. Under the policy scope for natural-language violations, this is a locale/language constraint that appears to be imposed without opt-in or justification.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The entire skill definition mandates Chinese-language persona, vocabulary substitutions, and stylistic rules such as replacing English business jargon with specific Chinese terms. This constitutes a language/locale constraint in natural-language instructions, but the file does not indicate that the user can choose another language or explicitly opt into Chinese-only behavior.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The entire skill definition is written as a Chinese-only persona and style guide, with no indication that the user can choose another language or locale. Under the stated policy, forcing a specific language without opt-in is a natural-language policy violation unless the constraint is clearly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The entire role definition is written as a prescriptive persona in Chinese and provides no indication that the user may choose another language. Under the stated policy, forcing a specific language without opt-in is a natural-language policy violation unless the locale restriction is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 82% confidence
Finding: The file is entirely framed as a Chinese-language brand-copy role and provides no indication that language should follow user preference or that Chinese is optional. Under the language/locale policy, forcing a specific language without opt-in is a natural-language policy concern.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The entire role specification is written as a Chinese-only persona/style guide and does not indicate that users may choose another language or that the skill is limited to a Chinese-language context. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The entire skill is written as prescriptive guidance for producing social-media copy in Chinese, including Chinese slang, phrasing, and stylistic constraints, but it does not indicate that language use is optional or region-specific. Under the policy, forcing a specific language or locale without user opt-in is a natural-language policy violation.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: This markdown role definition is entirely written as a prescriptive persona in Chinese and does not offer the user any language or locale choice. Under the policy criteria, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The entire skill definition is written as prescriptive guidance in Chinese and does not indicate that language is selectable based on user preference. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The entire skill is written as a role specification in Chinese and presents a fixed writing persona/style for speeches without any indication that users may choose another language or locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The document is written as mandatory guidance and states that the content applies to all roles, while all required rewriting behaviors and examples are specified exclusively in Chinese. This effectively forces a specific language/locale for output without user opt-in, which matches the language/locale policy violation criteria.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The README states “专为中文文本设计,” which sets a Chinese-language constraint without indicating user choice or opt-in. This is a natural-language locale restriction and is not accompanied by language-selection guidance.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.