ClawHub
Back to skill

Security audit

Ponytail Help

Security checks across malware telemetry and agentic risk

Overview

This is a small help-only skill that documents Ponytail commands and does not contain executable code or hidden data access.

This looks safe to install as a help/reference skill. Be aware that the documented phrase "normal mode" is a broad Ponytail deactivation alias, so prefer explicit commands like `/ponytail off` when changing modes.

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
90% confidence
Finding
The deactivation trigger includes the plain-language phrase "normal mode," which is likely to appear in ordinary conversation and can unintentionally disable the skill. In an agentic setting, broad trigger phrases are vulnerable to accidental invocation by users, tool output, or untrusted content, causing mode changes without clear intent.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal