ClawHub
Back to skill

Security audit

Cron Helper

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed cron scheduling helper; its command execution and recurring-task behavior are aligned with that purpose, though users should treat scheduled agent actions carefully.

Install only if you want your agent to create recurring tasks. Review each scheduled task before adding it, avoid broad tasks that access sensitive accounts or run open-ended commands, and periodically use the list/log/delete controls to audit what is active.

SkillSpector

Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
90% confidence
Finding
The manifest asserts that risky code was removed and safety improved, yet the skill still declares the `exec` tool, which enables command execution. This mismatch can create unjustified trust and reduce operator scrutiny around a capability that can run system commands, increasing the chance of unsafe task execution through the skill.

Session Persistence

Medium
Category
Rogue Agent
Content
# Cron Helper

Schedule recurring tasks for your AI agent without wrestling cron syntax.

## Features
Confidence
86% confidence
Finding
The skill is explicitly designed to create recurring tasks, which introduces persistence beyond the current session. In a skill with `exec` capability, persistent scheduled actions can repeatedly run commands or prompts over time, amplifying the impact of a mistaken, malicious, or overly broad scheduled task.

Static analysis

No suspicious patterns detected.