ClawHub
Back to skill

Security audit

Boxed fs

Security checks across malware telemetry and agentic risk

Overview

This skill is coherent for sandboxed file work, but it asks users to download and run an unpinned remote WASM component that can modify or delete files in mapped directories.

Install only if you are comfortable trusting a separately downloaded WASM binary from an unpinned source. Keep workDir and mapDirs as narrow as possible, avoid mapping sensitive folders, and review every write, overwrite, or remove command before allowing it.

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 condition is broad enough to activate on almost any file-manipulation request, including potentially sensitive or destructive actions. In an agent setting, ambiguous triggering can cause the skill to be selected without clear user intent or sufficient guardrails, increasing the chance of unsafe file access or modification.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill advertises destructive operations like remove-path and overwrite-capable write operations without any warning, confirmation requirement, or safety notes. This makes accidental data loss more likely, especially because the skill is framed as generally secure due to sandboxing, which may cause users or agents to underestimate deletion and overwrite risk.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.