ClawHub
Back to skill

Security audit

CP2K Cross-Code Input Studio

Security checks across malware telemetry and agentic risk

Overview

This skill is an offline CP2K input draft generator/converter with disclosed local Python helpers and no evidence of hidden data access or harmful behavior.

Install only if you are comfortable with bundled local Python scripts that read chemistry request/structure files and write generated draft outputs. Run them in a controlled project directory and review the scientific settings before using the generated inputs.

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 explicitly instructs the agent to read local reference files and use local scripts that normalize requests, render inputs, write output files, and perform format conversion, which implies file read, file write, and shell/code-execution capabilities despite no declared permissions. This mismatch is dangerous because it obscures the skill's actual execution surface, preventing policy enforcement and making it easier for a downstream agent to invoke filesystem or script actions without explicit approval boundaries.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.