ClawHub
Back to skill

Security audit

novel-multi-agent-skill

Security checks across malware telemetry and agentic risk

Overview

This appears to be a local novel-writing workflow skill, with the main caveat that it enforces strict editorial content rules and includes AI-style cleanup guidance.

Review the content redlines and AI-transparency wording before installing. This skill is suitable if you want a strict Chinese web-novel production workflow, but users publishing commercially or in disclosure-sensitive contexts should adjust the templates to represent AI assistance honestly.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Natural-Language Policy Violations

Medium
Confidence
94% confidence
Finding
This file imposes a mandatory China-specific censorship and localization regime on all agents and all novel projects, removing user choice over language, locale, and content policy. That creates a policy-manipulation vulnerability: the skill can force agents to suppress legitimate topics, rewrite real entities into aliases, and systematically bias outputs regardless of user intent or higher-level safety requirements.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The template explicitly instructs authors to 'remove AI traces,' which encourages concealment of AI involvement rather than transparent disclosure. In a content-generation workflow, this can facilitate deceptive attribution, undermine user trust, and create compliance or policy issues where AI assistance must be disclosed or approved.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.