Security audit

游戏技能

Security checks for vulnerabilities and agentic risk

Overview

This game-development skill is not clearly malicious, but it asks for broad read, write, command, API, and API-key authority without tight scoping to game-development workflows.

Install only if you are comfortable with a game-development assistant that may read and modify project files, run local commands, and use an API key. Keep it limited to trusted game project directories, review commands before execution, and avoid providing credentials or private project data unless the destination service is clear.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 93% confidence
Finding: The skill is presented as a game-development helper, but its documented capabilities broaden into generic file handling, API calls, and command execution. This scope mismatch increases the chance the agent will invoke powerful operations outside the user's expected task boundary, enabling abuse or unintended system actions under a misleading label.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The skill advertises command execution as a general capability without tying it to specific game-development workflows or safety constraints. Generic exec access materially raises the risk of arbitrary command execution, filesystem modification, toolchain abuse, or lateral misuse if the skill is triggered inappropriately.

Context-Inappropriate Capability

Medium

Confidence: 90% confidence
Finding: The skill claims standardized external API integration even though the manifest scope is game development rather than a defined API client. This unnecessary network capability can be abused for unreviewed outbound data transfer, use of untrusted services, or expansion of the skill into a generic automation bridge.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The activation guidance is broad and vague, mixing English and Chinese triggers such as using the skill whenever users need related functionality. Overbroad trigger conditions can cause the agent to invoke a high-privilege skill too often, including in contexts where file writes or command execution are unnecessary or unsafe.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The documentation describes file writing, API use, and command execution but does not provide a clear, prominent warning that these actions can alter the system or exfiltrate data. In a skill exposed through an agent, missing warnings reduce informed consent and make risky operations more likely to be performed implicitly.

Static analysis

No suspicious patterns detected.