ClawHub
Back to skill

Security audit

游戏

Security checks across malware telemetry and agentic risk

Overview

The skill is a game generator, but it asks for broad read/write/command authority and describes generic API and shell capabilities without clear boundaries.

Review this skill before installing if you do not want a game-generation assistant to have shell command and broad file authority. Use it only in a controlled workspace, approve package installs or commands explicitly, and avoid providing API keys unless a specific endpoint and purpose are 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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The skill claims generic file handling, API integration, and especially system command execution capabilities that are broader than necessary for a game-generation skill. This expands the attack surface significantly: a prompt-influenced agent could be induced to write arbitrary files, invoke shell commands, or interact with external services under the guise of generating a game.

Context-Inappropriate Capability

Medium
Confidence
89% confidence
Finding
The skill advertises external API use and includes network/authentication error handling without clearly tying those behaviors to a necessary, bounded workflow. In an agent setting, unjustified network capability can enable data exfiltration, unreviewed third-party requests, or secret misuse if prompts steer the skill toward arbitrary remote interactions.

Vague Triggers

Medium
Confidence
76% confidence
Finding
The activation and usage description is broad and permissive, telling the agent to invoke the skill with 'necessary input parameters' without clear trigger boundaries or safety constraints. Overbroad activation increases the chance the skill will be selected in contexts where its powerful tools—read, write, exec—are unnecessary, enabling prompt-driven misuse.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The markdown describes file writing, command execution, and external API use as product features but does not provide clear user-facing warnings about system impact, data exposure, or operational risk. In practice, this can normalize dangerous actions and cause agents or users to authorize powerful operations without understanding that files may be modified, commands run, or data transmitted externally.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.