ClawHub

is-it-true

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed critical-thinking and fact-checking prompt, with only a separate manual packaging helper as the notable non-runtime code.

Install this if you want a structured fact-checking and argument-analysis prompt. Avoid submitting confidential material unless you are comfortable with it being used in web searches, and only run the packaging script on skill directories you intend to package.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding
The skill appears to use file-reading capabilities without declaring corresponding permissions, which breaks the principle of least privilege and prevents users or the platform from accurately understanding what the skill can access. In this context, a skill presented as content-analysis logic has little obvious need to inspect local package files at runtime, so undeclared file access increases the risk of hidden data exposure or unauthorized local inspection.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The documented purpose is critical analysis of claims and web content, but the detected behavior includes package validation, metadata parsing, filesystem walking, and archive creation. That mismatch is dangerous because it can conceal non-obvious local file and packaging operations behind an innocuous description, undermining informed consent and creating opportunity for unintended file access or supply-chain style abuse.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal