ClawHub
Back to skill

Security audit

Azure Toolkit Free

Security checks across malware telemetry and agentic risk

Overview

This Azure management skill is openly about cloud resource operations, but it gives agents broad command execution and destructive Azure authority without enough safety scoping or confirmation guidance.

Review this skill carefully before installing. Use it only in a constrained Azure subscription or resource group, give any service principal the minimum permissions needed, avoid storing secrets in project files, and require the agent to show a preflight plan and get explicit approval before deleting, stopping, or changing network-access resources.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
86% confidence
Finding
The skill declares both write and exec capabilities, which can enable arbitrary local command execution beyond the stated free-tier Azure-management purpose. In an agent environment, overly broad execution permissions increase the blast radius of prompt injection, parameter abuse, or accidental destructive actions on the host and cloud resources.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The documentation explicitly includes destructive operations such as stopping and deleting resources without warning users about irreversible effects, downtime, or billing and data-loss implications. In a cloud-management skill, omission of such safeguards can lead to accidental service disruption or resource destruction when an agent follows user instructions too eagerly.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill documents use of Azure subscription IDs, tenant IDs, client IDs, and especially client secrets, but does not provide strong guidance on secret handling, storage, redaction, and non-disclosure in logs or outputs. In agent workflows with read/write/exec capability, this increases the risk of credential exposure, reuse, or accidental persistence in files, prompts, or execution logs.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.