ClawHub
Back to skill

Security audit

officecli

Security checks across malware telemetry and agentic risk

Overview

The skill is coherent for Office document work, but its install instructions tell users or agents to run remote shell scripts without verification or a warning.

Review the installer source and trust model before installing. Prefer a signed package, pinned release, checksum or signature verification, or manual inspection of the install script before running it. Once installed, use the tool only on documents you intend to modify and close resident/watch sessions when finished.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill instructs users to execute remote installer scripts by piping network content directly into bash or PowerShell, which creates an immediate code-execution path with no integrity verification, pinning, or review step. In an agent skill context, this is more dangerous than ordinary documentation because it can normalize unsafe installation behavior and potentially lead an automation-capable agent or user to run untrusted code from a third-party domain.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The installation section uses curl/irm piped directly into shell execution without any warning that this runs arbitrary remote code with the user's privileges. The absence of a security warning materially increases the risk of accidental compromise, especially in a skill that may be followed verbatim by users or automation systems seeking to satisfy document-editing tasks.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.