ClawHub
Back to skill

Security audit

openmath-lean-theorem

Security checks across malware telemetry and agentic risk

Overview

The skill is a disclosed Lean theorem workflow helper; its command execution and optional external skill installation are purpose-aligned and user-directed.

Install it only for the OpenMath Lean theorem workflow. Review the external leanprover/skills source before using npx or auto-install, prefer a project-local skills directory such as .codex/skills, and run the preflight only against the intended theorem workspace.

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
90% confidence
Finding
The skill instructs the agent to invoke shell commands, read environment variables, and potentially write files via auto-install workflows, but it does not declare corresponding permissions. This creates a capability transparency gap: a user or runner may approve or execute the skill without understanding that it can access local environment state and modify the filesystem, including by cloning or installing external content.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.