ClawHub
Back to skill

Security audit

多Agent协作小说创作系统

Security checks across malware telemetry and agentic risk

Overview

This is a coherent novel-writing assistant that creates local draft files, with broad activation wording but no hidden commands, credential access, network calls, or destructive behavior.

Install this if you want a Chinese novel-planning and drafting workflow. Use the step-by-step mode when you want confirmation between stages, and avoid giving sensitive personal or unpublished material unless you are comfortable with it being saved in the generated project files.

SkillSpector

By NVIDIA
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 (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger phrases are broad enough to match ordinary brainstorming or casual writing requests, which can cause this skill to activate when the user did not explicitly ask for a multi-agent novel-writing workflow. Over-broad activation increases the chance of unintended context capture, incorrect tool behavior, or surprising workflow execution, especially because the skill defaults to automatically running the full process.

Natural-Language Policy Violations

Medium
Confidence
83% confidence
Finding
The README is entirely Chinese-oriented and does not offer a language choice or explain any justified locale restriction, which can cause users or orchestrators to assume Chinese-only interaction regardless of user preference. This is primarily a safety and usability issue: it may lead to misrouting, misunderstood instructions, or exclusion of users, but it does not by itself introduce a direct security exploit.

Vague Triggers

High
Confidence
95% confidence
Finding
The skill is configured to trigger on very broad phrases such as vague requests to 'write down an idea' or 'help me think of a story,' which can capture many ordinary creative-assistance queries without clear user intent to invoke this specific multi-agent workflow. That increases the risk of unexpected activation, over-collection of user input, and unintended creation of structured project artifacts or long-form outputs the user did not explicitly request.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill defines a workflow that creates and stores multiple files, drafts, reviews, and a final manuscript, but it does not clearly warn the user in advance that their content may be persisted in a structured project directory. This can create privacy and consent issues, especially when users provide sensitive personal ideas, autobiographical material, or unpublished intellectual property expecting only transient assistance.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.