ClawHub
Back to skill

Security audit

challenge

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed local challenge workflow that reads and updates Gauntlet state files to quiz contributors, with no evidence of hidden or destructive behavior.

Before installing, expect the skill to read and update local `.gauntlet` challenge state and developer progress, and be aware that if a pending challenge exists it may treat your latest message as the answer. Use it in repositories where Gauntlet state tracking is intended.

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 (2)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger "testing contributor knowledge of the codebase" is broad enough that the skill may activate in contexts where a user is discussing testing, onboarding, or repository understanding without explicitly requesting this challenge workflow. Unintended activation is security-relevant here because the skill reads and writes local state files and can influence user interactions by evaluating prior messages as answers.

Missing User Warnings

Low
Confidence
81% confidence
Finding
The skill behavior includes reading `.gauntlet/knowledge.json`, checking pending state, and updating developer progress and streak, but the description does not warn that local files are accessed and modified. This reduces transparency and can surprise users, especially because prior messages may be evaluated and state may be persisted without an obvious disclosure at activation time.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.