ClawHub

Hekouwang Claude Skill Doctor Skill

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent local Skill-audit tool that reads a user-chosen skill directory, reports quality/security findings, and only edits files after user approval.

Install it if you want a local quality checker for Agent Skills. Use it on specific skill folders, not broad private workspaces, and review proposed edits before approving them. Do not keep real credentials inside skill bundles; despite its stated safety rule, the script may inspect ordinary text/source files with secret-like names if they are in the target directory.

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 (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The README advertises very broad natural-language trigger phrases such as '检查我的 skill', 'audit this skill', and any request about evaluating or optimizing an Agent Skill. In an agent ecosystem that auto-loads skills from user phrasing, overly broad triggers can cause unintended invocation on loosely related requests, expanding the skill’s influence and potentially exposing local files or causing unnecessary analysis steps.

Vague Triggers

Medium
Confidence
80% confidence
Finding
The trigger text includes a very broad condition: any request to evaluate, review, or optimize an Agent Skill's quality or structure should activate this skill. Overbroad activation can cause unintended invocation in contexts where another skill is more appropriate, increasing unnecessary file access and instruction interference from a large auditing prompt.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal