ClawHub

AI Company HR

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only AI-agent HR workflow skill; its broad permissions fit the stated orchestration purpose, but users should constrain triggers and external actions before relying on it.

Install only if you intend to use it for AI-agent HR governance. Narrow activation to explicit AI-company HR tasks, restrict network/API access to trusted systems, and require human approval plus audit records before any agent retirement, permission change, data deletion, or subagent delegation.

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 (2)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger list includes broad generic terms like "HR," "ethics," and common Chinese equivalents that can cause the skill to be invoked outside the user's intent. Because this skill has write, network, session, and subagent capabilities, accidental invocation could route sensitive HR-style tasks into a high-privilege workflow and expand the blast radius of benign prompts.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill declares file write, network API, sessions_send, and subagents permissions, but the description does not clearly warn users that invoking it may perform external actions or delegate work. This reduces informed consent and increases the risk that users trigger data movement, persistence, or multi-agent execution without realizing the operational and privacy implications.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal