ClawHub
Back to skill

Security audit

Create Agent App

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed TypeScript agent-app generator that requires user confirmation before making changes and includes safety and validation guidance.

Install this if you want Codex to help build or refactor TypeScript agent applications. Review the proposed architecture and brief-to-file mapping carefully before confirming, especially when package installs, file writes, network calls, credentials, databases, browser automation, or external API mutations are involved.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The skill enables implicit invocation while only providing a very broad description and default prompt, which can cause the agent to auto-select this powerful code-generation skill in contexts the user did not clearly intend. Because the skill can scaffold or refactor production-grade agent applications, unintended activation could lead to overbroad code generation, unsafe architectural changes, or misuse in sensitive environments without explicit user confirmation.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.