ClawHub

skillscope

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only skill recommender that clearly uses an external Skillscope API, with privacy considerations but no hidden code, persistence, or destructive behavior.

Install only if you are comfortable sending skill-recommendation queries to skillscope.cn. Do not include confidential task details, and inspect any recommended skill and install command before running it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The activation guidance is extremely broad: it triggers when a user asks for any skill for any task or merely describes a task that a skill could help with. This can cause over-invocation of the skill and unnecessary routing of user requests to a third-party recommendation service, increasing privacy exposure and the chance of irrelevant or premature tool use.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill instructs the agent to send the user's task and optional context fields to a remote API, but does not include a user-facing disclosure or consent step. Because task descriptions may contain sensitive business, personal, or operational information, silent transmission to an external service creates a real data leakage risk.

Natural-Language Policy Violations

Medium
Confidence
89% confidence
Finding
The documentation states that region is auto-inferred if omitted, meaning locale or geographic preferences may be derived and used without explicit user opt-in. This can expose location-related metadata and steer recommendations based on inferred attributes the user did not knowingly provide.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal