ClawHub
Back to skill

Security audit

Python Executor

Security checks across malware telemetry and agentic risk

Overview

This skill is a clearly disclosed remote Python runner with real data-sharing risks, but the artifacts match that purpose and do not show hidden or malicious behavior.

Install only if you trust inference.sh and intend to let an agent run Python code through that provider. Prefer manual install with checksum verification over piping an installer into a shell, review code before execution, and avoid sending secrets, private datasets, credentials, or confidential generated files unless you explicitly intend that data to be processed or returned by the service.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The trigger list includes broad phrases such as 'python', 'execute code', 'run script', and generic task categories that are likely to match many ordinary user requests. That can cause unintended invocation of a code-execution skill, increasing the chance an agent routes sensitive data or untrusted instructions into a powerful execution environment without clear user intent.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill prominently demonstrates outbound HTTP requests, web scraping, API access, and automatic return of files from outputs/ without adjacent warnings about data egress, third-party transmission, or artifact exposure. In a code-execution skill, this materially increases risk because users or upstream agents may send sensitive inputs to external services or unintentionally return generated files containing confidential data.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.