ClawHub

ClawGuard-Auditor

Security checks across malware telemetry and agentic risk

Overview

This is a local skill-auditing tool, but it overstates its security capabilities and describes auto-approval decisions that users should not rely on without review.

Install only if you want a heuristic local auditor and will review its output yourself. Do not treat its APPROVED or low-risk result as a definitive security decision, and run it only against the specific skill directory you intend to inspect.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
84% confidence
Finding
The skill claims enterprise-grade ML anomaly detection, supply-chain verification, and comprehensive audit capabilities, but the described or detected implementation appears to be largely heuristic and incomplete. This mismatch can create a false sense of security, causing users to trust approval decisions that are less rigorous than advertised.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger phrases are broad natural-language requests such as '审计这个 skill' and '安装前检查', which are plausible in normal conversation and can cause the skill to activate unintentionally. For a security-auditing skill, accidental invocation can expose local paths, read untrusted skill content automatically, or interfere with the user's intended workflow.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
| Classification | Score Range | Action |
|---------------|-------------|--------|
| **Trusted** | 80-100 | Auto-approve with standard logging |
| **Verified** | 60-79 | Approve with enhanced logging |
| **Unknown** | 40-59 | Manual review required |
| **Suspicious** | 20-39 | Deep audit required |
Confidence
76% confidence
Finding
Auto-approve

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
| Tier | Score Range | Color | Action |
|------|-------------|-------|--------|
| **TIER_0** | 0-10 | 🟢 GREEN | Auto-approve |
| **TIER_1** | 11-30 | 🟢 GREEN | Approve with logging |
| **TIER_2** | 31-50 | 🟡 YELLOW | Manual review |
| **TIER_3** | 51-70 | 🟠 ORANGE | Deep audit required |
Confidence
78% confidence
Finding
Auto-approve

VirusTotal

50/50 vendors flagged this skill as clean.

View on VirusTotal