ClawHub
Back to skill

Security audit

Azure智能体框架工具-专业版

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Azure agent orchestration guide with expected cloud, command, and file-write examples, but users should treat its deployment and credential-related steps as operationally sensitive.

Install only if you intend to let an agent help manage Azure AI Foundry resources. Review commands before execution, use staging first, keep real webhook URLs and credentials out of committed files, and ensure Azure credentials have least-privilege access.

SkillSpector

Vulnerability Patterns
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The skill states that secrets should not be hardcoded, yet its configuration examples embed operational webhook URLs directly in config/environment examples. While a webhook URL is not always a secret by itself, treating notification endpoints as harmless can normalize storing sensitive integration endpoints in files that may be committed, reused across tenants, or exposed in logs. In an enterprise orchestration skill that supports alerts and multi-tenant operations, this increases the chance of information leakage or unauthorized alert injection/spam.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill includes examples that run deployment commands, health checks, metric exports, and file writes without prominent safety gating or user-facing warnings immediately around those instructions. In a skill that declares tools including exec and write, broad examples can lead an agent to perform state-changing actions on local files or remote environments with insufficient confirmation, creating risk of accidental deployment, overwriting artifacts, or leaking operational data.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.