ClawHub
Back to skill

Security audit

Test Skill

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a purpose-aligned publishing helper, with file-upload risks users should understand before using it.

Before running the publisher, use it only from a clean package directory and check that no secrets, credentials, private notes, or unrelated files are present. Treat the upload bug as a quality issue that should be fixed, but the available evidence does not justify blocking the skill as malicious or requiring Review.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The helper function claims to upload file content but never defines or reads `content`, causing a runtime failure and making the upload path unreliable. In security-sensitive publishing code, broken upload logic can lead to unexpected behavior, poor validation, and operators misunderstanding what data is actually transmitted.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
This script uploads local files from a user-specified directory to a remote service, but it does not clearly warn the user which files will be transmitted beyond basic status output. In an agent-skill context, this is sensitive because skills often contain code, prompts, configuration, or embedded secrets, and accidental over-sharing can expose proprietary or confidential material.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.