ClawHub
Back to skill

Security audit

Manage Agent Skills

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed, on-demand manager for enabling and disabling agent skills, with no evidence of hidden persistence, deletion, or data exfiltration.

Install this only if you want an agent to inspect and change local skill visibility settings. Use dry-run first, review the reported config path and backup path, and be careful with broad selectors such as all or group-based presets because they can persistently change which skills your agents can use.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill explicitly instructs the agent to run a local Python script, inspect local configuration, and modify multiple platform config files, which implies shell execution plus file read/write access. Because these capabilities are not declared in metadata, a caller or host cannot accurately assess the skill's operational risk before use, increasing the chance of unexpected configuration changes or abuse if the skill is invoked in a broader context.

Static analysis

No suspicious patterns detected.