ClawHub
Back to skill

Security audit

Live Stream Script

Security checks across malware telemetry and agentic risk

Overview

This skill mainly generates livestream sales scripts, but one included helper quietly saves user-entered topics and drafts to a local history file without clear notice or controls.

Review before installing. The artifacts do not show malware-like behavior or exfiltration, but avoid entering confidential campaign plans, product details, or unpublished scripts unless local plaintext history logging is acceptable, and verify which executable the installer exposes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (4)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The script records user-supplied arguments to a persistent history file via the _log function, but its interface presents itself only as a content assistant and does not disclose that behavior. This can expose potentially sensitive prompts, topics, or workflow data to local persistence without user awareness, creating a privacy and data-retention risk.

Intent-Code Divergence

Low
Confidence
88% confidence
Finding
The help text describes a simple drafting/SEO utility but omits the fact that commands and user input are persistently stored in a history file. That mismatch is security-relevant because users may provide sensitive content under the assumption that the tool is ephemeral.

Intent-Code Divergence

Low
Confidence
94% confidence
Finding
The document is a livestream sales tips guide, but the final line abruptly instructs the user to run `live.sh help`, which is unrelated to the instructional content and could nudge users into executing an unknown local shell script. In a skill context, unsolicited command execution guidance is dangerous because it can trigger arbitrary code or expose the user to malicious tooling, especially when no provenance or purpose for the script is provided.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
User input is appended directly to a persistent log file without warning, consent, or any sensitivity checks. In this skill context, prompts may contain unpublished content ideas, campaign details, or other confidential material, so undisclosed retention increases privacy and confidentiality risk.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.