ClawHub
Back to skill

Security audit

Kling

Security checks across malware telemetry and agentic risk

Overview

This skill is a scoped guidance skill for creating Kling video briefs and includes clear consent, disclosure, and human-review limits.

Install only if you are comfortable using Kling as an external video-generation provider. Keep consent records for motion or avatar references, avoid uploading sensitive brand or personal material unless approved, and keep the stated human review and AI-disclosure steps in place before publishing.

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 (1)

Vague Triggers

Low
Confidence
76% confidence
Finding
The skill description uses broad routing language such as when to use Kling and its relationship to sibling skills, but it does not define clear invocation boundaries or exclusion rules. In an agentic router, ambiguous trigger scope can cause the wrong skill to be selected for tasks involving audio, editing, publishing, or identity-sensitive motion transfer, increasing the chance of policy bypass or unsafe execution paths.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.