ClawHub
Back to skill

Security audit

vault-data-governance

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed data-governance cleanup skill, but it grants very broad authority to move files across team workspaces and push repository changes without clear review controls.

Review this skill carefully before installing. It appears intended for legitimate vault cleanup, not deception or exfiltration, but it should be limited to explicit operator-run maintenance windows with dry-run output, path allowlists, diff review, and approval before any move, commit, or push.

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 (2)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger conditions are extremely broad ('after any agent completes a batch', manual audit/optimization, multiple workflow milestones), making it easy for this skill to run outside a narrowly controlled maintenance window. In context, the skill is authorized to move files across many shared directories and team workspaces, so accidental invocation can cause widespread unintended archival or disruption of active work.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The workflow performs bulk state-changing operations—moving files, staging all changes, committing, and pushing—without any explicit safety checks, user approval, branch isolation, or warning about repository-wide impact. Because the skill applies across all agents, task directories, and team workspaces, a mistaken or overbroad run could propagate destructive organizational changes into version control and shared remote history.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.