Security audit

Azure Global Search

Security checks for vulnerabilities and agentic risk

Overview

This skill is an Azure documentation and product lookup guide; it includes some cloud-provisioning examples, but they are disclosed as examples and the skill itself does not execute Azure changes or collect credentials.

Installers should treat this as a documentation helper, not an Azure automation tool. Review any generated CLI, SDK, REST, or deployment examples before running them, use least-privilege Azure roles instead of broad Contributor access where possible, and verify current pricing and product limits on official Microsoft sources.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (6)

Tool Parameter Abuse

High

Category: Tool Misuse
Confidence: 80% confidence
Finding: Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
Content: Scanner excerpt · examples.md (reported line 258)May include surrounding context.
md
| Get VM | GET /subscriptions/{sub}/resourceGroups/{rg}/providers/Microsoft.Compute/virtualMachines/{name} | Get VM details | | Start VM | POST /subscriptions/{sub}/resourceGroups/{rg}/providers/Microsoft.Compute/virtualMachines/{name}/start | Start a VM | | Stop VM | POST /subscriptions/{sub}/resourceGroups/{rg}/providers/Microsoft.Compute/virtualMachines/{name}/powerOff | Stop a VM | | Delete VM | DELETE /subscriptions/{sub}/resourceGroups/{rg}/providers/Microsoft.Compute/virtualMachines/{name} | Delete a VM | ### SDK Support | SDK支持 - **Python**: `azure-mgmt-compute`

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The description-level activation condition is broad enough that the skill may be invoked for many generic mentions of Azure, even when the user is not asking for Azure product discovery. This can cause unintended routing, irrelevant tool use, and interference with more appropriate skills, though it does not directly expose secrets or enable code execution.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The trigger condition explicitly says that any mention of "Azure" or "Microsoft Azure" can activate the skill, which is overly permissive. In a multi-skill environment this can lead to misfires, unnecessary web access, and reduced reliability by hijacking conversations that only mention Azure incidentally.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The document labels itself as a product query skill while later demonstrating provisioning and deployment workflows. This inconsistency undermines trust boundaries: agents or reviewers may classify the skill as read-only when it effectively normalizes operational guidance that can be turned into write actions against Azure resources.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill is described as a product-information lookup tool, but these examples teach operational Azure resource creation and code deployment. That scope expansion is dangerous because an agent invoking this skill could be induced to generate or facilitate real provisioning actions, increasing the chance of unintended cloud changes, cost incurrence, or deployment of insecure infrastructure.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The examples extend the skill from information retrieval into architecture recommendation and solution design, which exceeds the manifest's narrow query-focused description. While less severe than direct provisioning, this mismatch can still mislead orchestrators or users into trusting the skill for advisory decisions outside its declared scope, potentially resulting in unsafe or costly downstream actions.
Content: No source excerpt is available for this finding.

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical

Code: suspicious.exposed_secret_literal
Location: examples.md:639