ClawHub
Back to skill

Security audit

YiHui CONTEXT MODE

Security checks across malware telemetry and agentic risk

Overview

This context-management skill appears purpose-aligned, but it asks the agent to route broad file, shell, web, and persistent indexing activity through an MCP server with insufficient scoping in the submitted skill text.

Review before installing. Use this only in workspaces where you are comfortable granting an MCP server broad local execution, file-reading/indexing, and web-fetching ability. Pin and verify the npm package, configure deny rules for secrets and destructive commands, avoid indexing private logs or credential files, and confirm how to purge its stored SQLite/FTS data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger set is broad enough to match many normal requests such as project analysis, searching, or checking updates, which can cause the skill to activate unexpectedly. Because this skill routes file, shell, and web content through its own MCP server, overbroad activation increases the chance of unintended data access, external fetching, or altered agent behavior without clear user intent.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.