Back to skill

Security audit

文件

Security checks for vulnerabilities and agentic risk

Overview

This file-organizer skill is not malicious, but it asks for broad file and command authority without enough limits or safeguards for duplicate cleanup and other file changes.

Install only if you are comfortable granting the agent authority over the target folders. Use it on a test folder first, ask for a preview before any move/delete operation, and avoid providing API keys or sensitive directories unless the skill clearly explains what data will be sent and changed.

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill is a file organizer, but the documentation grants external API/network capability via an LLM API dependency and API key configuration without showing why network access is necessary for local file classification. Unnecessary network capability expands the attack surface and creates risk of data exfiltration, especially if file contents or metadata are sent off-host during processing.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The documentation advertises command execution for a file-classification skill, which is broader than needed for organizing files and can enable arbitrary system actions if the agent maps user input into exec calls. Even though the text mentions a sandbox and whitelist, no concrete restrictions are defined in the skill, so this capability materially increases the chance of destructive file operations or host compromise.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill claims it can clean duplicate files and perform write operations, but it does not provide a prominent warning about deletion, overwrite, or other irreversible changes. In a file-management context, this is dangerous because users may trigger destructive actions on valuable data without a dry-run, backup, or confirmation step.

Content

No source excerpt is available for this finding.