ClawHub
Back to skill

Security audit

Deduplicator

Security checks across malware telemetry and agentic risk

Overview

This skill only provides session-level deduplication instructions and has no executable code or hidden access.

Install this only if you want the agent to suppress repeated list items or links during a conversation. Be aware that repeated results may be omitted, and use the documented reset command when you want deduplication to start over.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The usage guidance is broad enough that the skill may activate on many list-processing requests without clear boundaries or explicit invocation. In an agent setting, overly broad activation can cause unintended suppression of items, hide relevant results, or interfere with downstream reasoning by silently skipping content the user did not intend to deduplicate.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.