ClawHub
Back to skill

Security audit

Skill Compliance

Security checks across malware telemetry and agentic risk

Overview

This is a local, user-run compliance checker that reads skill files and optionally writes a report, with no hidden network, credential, persistence, or destructive behavior found.

Install only if you are comfortable running a local checker over the skill directories you select. Use the documented --dir and --output options carefully, since it can read files under the target skill and write a report to the path you provide.

SkillSpector

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

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill declares `environment: local` and `dependencies: none`, but its documented usage invokes a Python script against arbitrary directories and the static analyzer detected file read, file write, and shell capabilities with no corresponding permission declaration. This mismatch is dangerous because reviewers or platforms may underestimate what the skill can access or modify locally, reducing transparency and weakening sandboxing or trust decisions.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.