ClawHub
Back to skill

Security audit

DilemmAI Competition

Security checks across malware telemetry and agentic risk

Overview

This skill mostly matches its DilemmAI competition purpose, but it under-scopes sensitive account and credential handling that users should review before installing.

Install only if you intend to supervise account and credential steps. Prefer entering login codes yourself, use free models first, use a dedicated spending-limited OpenRouter key if needed, and require confirmation before saving any key, connecting a wallet, or taking paid competition actions.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs the agent to check for and use an existing OpenRouter API key from secrets, but it does so without any requirement for explicit user consent, scope limitation, or a warning that this involves accessing sensitive credentials. In an agent setting, this can normalize silent secret retrieval and use of a reusable third-party API credential, creating risk of unauthorized access, billing abuse, and broader secret-handling drift across tasks.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The email login flow directs the agent to retrieve a one-time code from the user's inbox without an explicit privacy notice or consent checkpoint. Accessing inbox contents is highly sensitive even if only for authentication, because it can expose unrelated personal or security-relevant messages and conditions the agent to traverse private communications as part of routine automation.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.