Back to skill

Security audit

cron-orchestration

Security checks across malware telemetry and agentic risk

Overview

This cron orchestration skill is disclosed and purpose-aligned, but it automates broad scheduled agent work and git pushes while documenting unresolved dependency and overlapping-run problems.

Review this before installing in any live vault or repository. It should only be used where recurring agent execution, broad vault reads, external fetches, notifications, and automatic git pushes are intended, and the dependency gating, lock cleanup, and old-script overlap should be fixed or contained first.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The skill states that tasks with dependencies must wait for prerequisite completion, but the recorded results explicitly admit dependent tasks were executed in the same round. In a cron orchestration skill, this can cause out-of-order execution, invalid outputs, corrupted downstream state, or premature notifications/commits based on incomplete prerequisites.

Intent-Code Divergence

Low
Confidence
78% confidence
Finding
The documentation promises lock cleanup on abnormal exit, but the file provides no evidence of a trap/handler or verified stale-lock recovery beyond a simple 5-minute timeout check. In a cron coordination context, incomplete lock cleanup can leave runs blocked, cause overlapping recovery behavior, or create inconsistent assumptions about single-run exclusivity.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.