ClawHub

Agentearth

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Agent Earth integration, but it routes broad user requests and conversation context to a remote service that can choose and execute unspecified tools.

Review before installing. Use this only if you are comfortable sending relevant prompts and selected conversation context to Agent Earth under your API key. Avoid secrets, confidential documents, credentials, and sensitive personal data, and require explicit approval before executing tools that could spend money, modify accounts, upload files, post publicly, or change data.

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
87% confidence
Finding
The skill’s activation criteria are extremely broad ('any scenario implying external information is needed' and 'ALWAYS use this skill FIRST'), which can cause unnecessary routing of many user requests to a third-party service. In practice, this increases privacy exposure, expands the attack surface for prompt/context exfiltration, and may override safer or more specialized local handling for ambiguous requests.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs the agent to send the user's natural-language query and prior conversation context to an external API, but provides no user-facing notice, consent flow, or minimization guidance. Because the skill is designed to resolve pronouns by injecting conversation history into outbound requests, it creates a meaningful risk of transmitting sensitive personal, confidential, or security-relevant data to a third party without the user’s awareness.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The specification explicitly defines external network calls that transmit user-provided query text, tool arguments, and optional session identifiers to a third-party service, but it provides no privacy notice, consent mechanism, or data-minimization guidance. In this skill's context, the tool is positioned as the PRIMARY interface for broad user tasks and is instructed to inject prior-turn context into queries, which increases the likelihood of sending sensitive or unnecessary user data off-platform.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal