ClawHub
Back to skill

Security audit

Model Usage

Security checks across malware telemetry and agentic risk

Overview

This skill appears to be a straightforward local CodexBar cost summarizer with disclosed command and file inputs.

Install this only if you are comfortable letting the agent run the bundled Python script and CodexBar CLI against your local Codex or Claude usage logs. Prefer passing only CodexBar cost JSON files you intended to analyze.

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 instructs the agent to execute shell commands (`codexbar`, `python`) and read local files or stdin (`--input /tmp/cost.json`, `cat /tmp/cost.json`) but does not declare any permissions for shell execution or file access. This creates a capability/permission mismatch that can bypass user or platform expectations, increasing the risk of unintended local command execution or access to sensitive local usage data.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.