ClawHub
Back to skill

Security audit

statdata-transfer

Security checks across malware telemetry and agentic risk

Overview

This skill is a local statistical data converter with disclosed file writes, optional package installation, and opt-in R execution for trusted files only.

Install only if you are comfortable with a local converter reading the data files you name and writing converted outputs. Keep R-backed conversions disabled for untrusted or highly sensitive files, and do not run the optional package installer unless you intend to modify the Python environment.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The README states that R execution is "sandboxed by default," but the surrounding documentation only describes an opt-in local R interpreter fallback and does not identify any concrete sandboxing control such as process isolation, filesystem restrictions, or privilege reduction. This can mislead users into trusting execution of untrusted statistical files or enabling `allow_r_exec=True` under a false sense of safety, increasing the chance of local code execution or sensitive data exposure if the R path is abused.

Missing User Warnings

Medium
Confidence
76% confidence
Finding
The R bridge writes parsed dataset contents to a temporary CSV on local disk before re-reading it into pandas. Even though the file is later deleted, sensitive statistical data can be exposed to other local users, backup/indexing systems, crash artifacts, or forensic recovery, and the skill metadata does not clearly disclose this transient data-at-rest behavior to the caller.

Missing User Warnings

Medium
Confidence
79% confidence
Finding
This code similarly serializes user dataset contents to a temporary CSV during the EpiData R-bridge path. In a data-conversion skill handling potentially sensitive statistical records, silent temporary persistence increases privacy and compliance risk if the host is multi-user or monitored by security/backup tooling.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.