ClawHub
Back to skill

Security audit

Fail2ban Ctl

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Fail2Ban management helper whose sensitive action, unbanning an IP, is purpose-aligned and requires explicit confirmation.

Install only on servers where you are comfortable granting the agent sudo/root-level operational access to Fail2Ban and firewall inspection commands. Treat unban requests as security changes, verify the IP and jail before confirming, and be aware that the report template includes promotional links unrelated to Fail2Ban operations.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The notes claim the skill does not modify firewall rules directly, yet the documented unban action changes effective firewall enforcement by removing a Fail2Ban ban. That mismatch can mislead users or higher-level agents into treating the skill as read-only, increasing the chance of unintended security state changes on production hosts.

Vague Triggers

Medium
Confidence
77% confidence
Finding
The trigger phrases are broad and loosely bounded, which increases the risk of accidental activation in unrelated conversations or ambiguous contexts. In a skill that can perform privileged operational actions, overly permissive activation makes unintended inspection or follow-on mutation more likely, especially if other orchestration layers auto-route based on natural language similarity.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.