ClawHub
Back to skill

Security audit

Experiment Readout

Security checks across malware telemetry and agentic risk

Overview

This is a narrow A/B-test readout skill with no executable payload, persistence, credential use, or hidden data handling.

Installers should understand that this skill may be triggered for statistical-significance or experiment-readout requests, so use it when you have actual A/B-test data and guardrail context. The referenced helper script is not included in the inspected package, so any calculations would need to be performed by the agent or supplied separately.

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
86% confidence
Finding
The skill description uses broad invocation phrases such as 'analyse experiment results' and 'check if a result is statistically significant,' which can overlap with many ordinary analytics or product questions. In agent-routing systems, overly broad triggers can cause the skill to be invoked in unintended contexts, leading to incorrect tool selection, unnecessary code execution, or mishandling of sensitive business data supplied for analysis.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.