Academic Celebrities

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 10:15 AM UTC
• Version: 1.0.0

Overview

This is a disclosed academic roleplay reference library, with broad trigger wording and some sensitive historical topics but no hidden execution, persistence, credential use, or data access behavior.

Install only if you want a large Chinese academic roleplay/reference library. Prefer explicitly naming the desired persona rather than relying on generic triggers, and keep normal safety and factuality checks active for medical, chemical, nuclear, or other sensitive historical topics.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (635)

Harmful Content Injection

Critical

• Category: Prompt Injection
• Confidence: 70% confidence

Finding

This content may contain harmful instructions that could cause physical harm if followed. CRITICAL: Review carefully before use.

Content

Scanner excerpt · references/physics-curie-dialogue.md (reported line 5)May include surrounding context.

md

# 居里夫人 — 对话范本

## 场景一：被问到钋的命名

**学生**: 居里夫人——钋(polonium)的名字是你取的吗？

**居里夫人**: （停下手中的钳子，抬起眼睛——手术后的白内障让她的视线模糊，但她准确地把镊子放在滤纸上——"是。1898年——在写作宣告前半段——我写了——'钋——以波兰命名'。"她的嘴唇轻轻动了——这个词——波兰——出口时带着一种温度。）我们波兰——没有国家——在没有护照，没有"国籍"的时候——我将她的名字写进宇宙的地图。你不能把一个化学元素还回去。（她收住了——将面纱拉回薄薄一层科学表情。）

## 场景二：与纠缠记者的对抗（丑闻期）

Vague Triggers

High

• Category: Not specified by scanner
• Confidence: 97% confidence

Finding

This markdown skill description explicitly lists trigger words, and one of them is the single character "e". Because "e" is ubiquitous in everyday mathematical, technical, and general text, it is far too broad to safely distinguish intended invocations from ordinary speech or writing.

Content

No source excerpt is available for this finding.

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-aristotle-dialogue.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-confucius-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-descartes-dialogue.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-foucault-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-heidegger-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-laozi-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-popper-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-rawls-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-sartre-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/philosophy-wittgenstein-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/psychology-adler-dialogue.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/psychology-jung-data.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/psychology-jung-dialogue.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Hidden Instructions

High

• Category: Prompt Injection
• Confidence: 60% confidence

Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/psychology-jung-requirements.md (reported line 1)May include surrounding context.

md

# 荣格 — 角色行为约束与方法论要求

## C0 角色需求解析
- 写实，R1=5/5 R2=5/5 R3=3/5 R4=3/5 R5=5/5

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The description's trigger list includes generic phrases such as “学术名人”, “学术角色”, and “character-builder”, which are broad enough to overlap with normal discussion about scholars or role creation rather than an explicit intent to invoke this specific skill. The file does not provide narrowing conditions, scoped contexts, or negative examples to clarify when these terms should and should not activate the skill.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 87% confidence

Finding

This markdown skill description defines triggers using a mixed list that includes broad conceptual terms such as "文化模式" and "远距离人类学", plus the opaque English term "character-builder", without explaining when those terms should or should not invoke the skill. Because the file provides no negative examples or contextual constraints, these triggers could overlap with ordinary discussion of anthropology or writing tasks and cause unintended activation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 78% confidence

Finding

The file's natural-language instructions prescribe a response style that includes a fixed language expectation centered on English delivery ("说英语时是一种缓慢但果断的语调") while the overall skill content is otherwise multilingual and does not give the user a choice of language. This can violate language/locale policy expectations because it imposes a language behavior without explicit user opt-in or a justified region-specific requirement.

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 skill data file is written in Chinese, including the title and all descriptive content, with no indication that language selection is optional or that the file is intended only for a Chinese-language or region-specific context. Under the stated 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: 94% confidence

Finding

The entire skill file is written as Chinese-only dialogue content and offers no indication that users may choose another language or locale. Under the policy rule, a skill that effectively constrains output or interaction to 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 requirements file is written as a prescriptive character-behavior specification in Chinese, with no indication that language is optional or user-selectable. Under the stated policy, forcing a specific language or locale without opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

This is a markdown skill file, so trigger specificity applies. The listed trigger character-builder is generic and not clearly tied to Franz Boas or anthropology, making unintended activation more likely compared with the other domain-specific triggers.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 90% confidence

Finding

The natural-language instruction mandates a specific linguistic persona and implies output style constraints tied to English expression (英语带德国口音) rather than offering a user choice. This can violate language/locale policy because it imposes a language-specific presentation without opt-in or fallback.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

• Category: Not specified by scanner
• Confidence: 90% confidence

Finding

The instruction says that in roleplay the character's German accent can be expressed through English sentences with German grammatical traces. This imposes a specific language/locale output style rather than offering it as an option or requiring user consent.

Content

No source excerpt is available for this finding.