ClawHub
Back to skill

Security audit

驱动

Security checks for vulnerabilities and agentic risk

Overview

This brand-design skill is not malicious, but it asks for broad read, write, and command execution authority that is not well scoped to logo generation.

Review before installing. Use this only if you are comfortable granting the agent command execution and file read/write access, and restrict it to explicit user-requested brand-design actions such as generating documents or running a known CellCog command. Avoid giving it access to sensitive workspaces, credentials, or broad filesystem areas unless the publisher narrows the permissions and command scope.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (3)

Context-Inappropriate Capability

High
Confidence
95% confidence
Finding
This skill is advertised as a logo/brand identity generator, yet it declares command execution, file operations, and generic environment-management capabilities that materially exceed the stated purpose. That capability expansion increases the chance an agent will execute shell commands or perform system actions unrelated to design work, creating unnecessary attack surface for prompt injection, data access, or unintended system modification.

Intent-Code Divergence

Medium
Confidence
88% confidence
Finding
The file claims to be a pure Markdown, natural-language-driven skill while simultaneously documenting executable commands, API usage, file writes, and system command execution. This mismatch can mislead reviewers, users, or agent runtimes about the skill's true privileges and behavior, increasing the risk that powerful actions are enabled without appropriate scrutiny.

Description-Behavior Mismatch

Medium
Confidence
85% confidence
Finding
The manifest presents the skill as a focused brand-design tool, but the documented behavior expands into broad automation, file handling, API integration, and command execution. This scope mismatch is dangerous because users and policy systems may approve the skill under a low-risk creative pretext while it actually enables more powerful and potentially unsafe operations.

Static analysis

No suspicious patterns detected.