ClawHub

Ruflo

Security checks across malware telemetry and agentic risk

Overview

This does not look malicious, but it needs Review because it advertises persistent memory and federated agent communication without clear consent or data-boundary guardrails.

Install only if you are comfortable with a skill that may retain context and coordinate across agents or systems. Before use, require explicit confirmation for memory, federation, scripts, plugins, or sharing with other agents, and avoid sensitive prompts unless retention and data-boundary controls are documented.

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
93% confidence
Finding
The activation guidance is overly broad: 'when the user asks about multi-agent coordination, swarm orchestration, or agent memory' can match many ordinary requests and cause this powerful skill to be invoked without clear user intent. In a skill that references scripts, plugins, federated communications, and persistent memory, over-activation increases the chance of unintended cross-boundary behavior, unsafe tool use, or disclosure of stateful data.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill description advertises 'self-learning memory' and 'federated communications' but does not warn users that data may persist over time or potentially cross machine, team, or trust boundaries. In this context, the omission is meaningful because users may provide sensitive prompts assuming local, single-session handling when the skill is designed for persistent and federated operation.

VirusTotal

62/62 vendors flagged this skill as clean.

View on VirusTotal