Back to skill

Security audit

Cloud

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only cloud storage help skill with some sloppy boilerplate, but no artifact-backed malicious or high-impact behavior.

Before installing, treat this as a lightweight cloud-storage advice skill. Be aware that its trigger phrases and setup notes are broad and inconsistent, so review its advice before acting on account, sharing, backup, or deletion decisions.

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
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
99% confidence
Finding

The listed keywords—sync, cloud, organize, choose, share—are extremely vague and appear in many benign conversations. In an agent routing context, such vague triggers can cause accidental invocation, creating prompt-routing confusion and potentially exposing user context or causing the wrong skill to influence sensitive file-management decisions.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The front matter declares only a read tool, but later describes the skill as requiring or supporting exec/command-line execution. This mismatch is dangerous because it obscures the real capability expectations of the skill and could cause an agent or reviewer to underestimate operational risk, especially if the skill later evolves to invoke commands or assumes elevated permissions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The activation phrases are broad everyday language such as cloud storage, share folder, and where are my files, which can cause the skill to trigger in unrelated contexts. Overbroad triggering increases the chance that the skill is invoked for tasks outside its intended scope, potentially steering users into inappropriate advice or unauthorized handling of personal cloud-storage matters.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The documentation gives contradictory statements about API key requirements, which can mislead an agent or user about the trust boundary and setup assumptions for the skill. While this does not directly enable code execution, inconsistent security-relevant setup guidance increases the chance of unsafe operation, misconfiguration, or users providing secrets unnecessarily.

Content

No source excerpt is available for this finding.