Back to skill

Security audit

Academic Thesis Review

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed local thesis-review workflow; the main caution is that it saves extracted thesis text and review notes on disk.

Before installing, confirm you are comfortable with the skill saving extracted thesis text, working notes, and review results in the working directory. Use it in a dedicated folder for sensitive or unpublished theses, review generated artifacts before sharing them, and prefer an explicit invocation such as asking to use this thesis-review skill on a specific .docx file.

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
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
Findings (16)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The README states that the skill is designed to generate revision feedback in Chinese, which imposes a language constraint by default. Under the policy, forcing a specific language is a violation unless the user is given a language choice or the constraint is explicitly justified as region-specific.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
85% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.md (reported line 77)May include surrounding context.

Personal install:

bash
mkdir -p ~/.claude/skills/academic-thesis-review
cp -r . ~/.claude/skills/academic-thesis-review

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
85% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.zh-CN.md (reported line 83)May include surrounding context.

Personal install:

bash
mkdir -p ~/.claude/skills/academic-thesis-review
cp -r . ~/.claude/skills/academic-thesis-review

Session Persistence

Medium
Category
Rogue Agent
Confidence
84% confidence
Finding

The README encourages persistent installation into user-level or project-level skill directories, which means the behavior and prompts remain available across sessions. In the context of a skill that processes sensitive thesis documents and writes artifacts to disk, persistent installation increases the chance of repeated unintended use, lingering access to derived content, and longer retention of risky defaults.

Content

Scanner excerpt · README.md (reported line 77)May include surrounding context.

Personal install:

bash
mkdir -p ~/.claude/skills/academic-thesis-review
cp -r . ~/.claude/skills/academic-thesis-review

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
85% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.md (reported line 84)May include surrounding context.

Project install:

bash
mkdir -p .claude/skills/academic-thesis-review
cp -r . .claude/skills/academic-thesis-review

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
85% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.zh-CN.md (reported line 90)May include surrounding context.

Project install:

bash
mkdir -p .claude/skills/academic-thesis-review
cp -r . .claude/skills/academic-thesis-review

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The suggested trigger phrases include very broad everyday requests such as '帮我看看论文', which can cause accidental or overly permissive invocation in agent systems that route skills by fuzzy phrase matching. In a skill ecosystem, broad triggers can expose sensitive thesis content to this workflow unintentionally and activate file-writing behavior without the user's explicit selection of this specific skill.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README states that the skill writes review_results.md and thesis-derived artifacts to disk, but it does not warn users that potentially sensitive academic content will be persisted locally. Because theses may contain unpublished research, personal information, or proprietary organizational data, silent persistence increases confidentiality and retention risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

SQP-3 适用于所有文件类型。README 明确规定“输出风格为……中文修改意见”,属于语言/locale 约束,但文档未说明可切换语言、征得用户选择,或给出必须固定中文的合规性理由。

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · README.zh-CN.md (reported line 83)May include surrounding context.

个人目录安装:

bash
mkdir -p ~/.claude/skills/academic-thesis-review
cp -r . ~/.claude/skills/academic-thesis-review

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · README.zh-CN.md (reported line 83)May include surrounding context.

个人目录安装:

bash
mkdir -p ~/.claude/skills/academic-thesis-review
cp -r . ~/.claude/skills/academic-thesis-review

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

该 README 是 markdown 文件,适用 SQP-1。触发语句列表中的“审阅论文”“帮我看看论文”“修改后再看看”表述较宽泛,尤其后两者接近日常对话,且未给出排除条件或更明确的上下文约束,容易与一般求助或普通文档审阅请求重叠。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The description says the skill produces revision comments in Chinese, and the output specification later requires "Language: Chinese." This imposes a fixed language policy without opt-in or an explicit justification that the skill must operate only in Chinese.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases include very broad, everyday requests such as '帮我看看论文' and '论文有什么问题', which can match normal conversational input and cause the skill to activate when the user did not explicitly request this specialized workflow. Because the skill performs file-oriented, multi-step review behavior and can read/write local artifacts, unintended activation increases the risk of surprising actions, privacy exposure, and incorrect task routing.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

该 README 是 markdown 文件,适用 SQP-2。文档说明会在工作目录生成 review_results.md 并自动创建 review_artifacts/ 目录,但未明确提醒用户这会写入本地文件、可能包含论文内容提取结果,从而影响工作目录和本地数据留存。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The manifest description states the skill is for review of 'Chinese management-oriented master's theses,' which embeds a language/locale-specific constraint in the skill definition. Because the file does not indicate that this limitation is optional or user-selectable, it may conflict with the policy against forcing a specific language or locale without opt-in.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.