ClawHub

Verified Task

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed verification gate that tells agents to pause unless outputs pass review, without hidden data access or execution behavior.

Install this if you want agents to add an explicit verification checkpoint before important actions. Expect it to stop or delay workflows until outputs pass or a human override is recorded, and avoid sending secrets or sensitive content to any optional external verification service.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (2)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The 'When to Use This' section frames activation very broadly across common activities such as posting content, making decisions, and especially the catch-all statement 'If the output matters, verify it before proceeding.' For a markdown skill description, this lacks clear trigger boundaries or exclusion conditions and could cause unintended invocation in many ordinary workflows.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
* prevent incorrect execution
* enforce task correctness
* keep workflows on track during autonomous runs
* ensure important actions do not proceed without verification

---
Confidence
75% confidence
Finding
without verification

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal