ClawHub
Back to skill

Security audit

翰林学士

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward academic proofreading tool that sends user-provided text to DeepSeek as disclosed, with privacy considerations users should understand before use.

Install only if you are comfortable sending the text you ask it to polish to DeepSeek. Avoid using it for confidential manuscripts, regulated data, student records, or proprietary research unless your organization permits that third-party processing.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill instructs the agent to call an external DeepSeek API and references code capabilities such as network access, but no permissions are explicitly declared to make that behavior visible and reviewable. This weakens security transparency and governance, making it easier for data exfiltration or unexpected outbound requests to occur without clear user/admin awareness.

Vague Triggers

Medium
Confidence
72% confidence
Finding
The trigger conditions are broad enough to activate on generic requests like '改一下这段' or 'proofread,' which may capture non-academic or sensitive text unintentionally. In combination with the external API workflow, over-triggering increases the chance that unrelated private content is sent to a third party without the user clearly intending to invoke this skill.

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill directs submission of user text to an external DeepSeek API but does not provide a user-facing warning or obtain informed consent. Because users may paste unpublished papers, research data, or personal information, silent third-party transmission creates a significant confidentiality and compliance risk.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill sends user-supplied paper text to a third-party API, which is a real privacy and data-governance risk because academic drafts may contain unpublished research, proprietary information, or personal data. While external processing is the intended function of the skill, there is no explicit consent flow, warning, or data-handling notice before transmission.

External Transmission

Medium
Category
Data Exfiltration
Content
}

    try:
        response = requests.post(
            DEEPSEEK_API_URL, headers=headers, json=payload, timeout=60
        )
        if response.status_code >= 400:
Confidence
95% confidence
Finding
requests.post( DEEPSEEK_API_URL, headers=headers, json=

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.