ClawHub
Back to skill

Security audit

Hetzner

Security checks across malware telemetry and agentic risk

Overview

This Hetzner operations skill is coherent and disclosed, with local inventory writes and cloud-operation guidance that fit its stated purpose.

Install only if you want the agent to maintain local Hetzner inventory and memory files under ~/Clawic/data/ and to work with your existing hcloud/API context. Review any proposed delete, rebuild, resize, detach, or cancellation action before approving it, and keep actual tokens, passwords, keys, and rescue credentials in a secret manager rather than in notes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The template explicitly permits writing durable state with 'No permission needed' and defines many automatic write targets under shared paths. In an agent skill, this broad write authority can cause unauthorized persistence, cross-skill data tampering, or unintended modification of shared inventories without an explicit user confirmation boundary.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The split procedure instructs the agent to move and delete content from memory.md in the same turn and to 'never leave a copy behind,' but does not require backup, preview, or user approval. That creates a real integrity risk: a malformed transform, parser bug, or adversarial prompt could erase or relocate user data irreversibly.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
memory-template.md:60