ClawHub
Back to skill

Security audit

MiniMax XLSX Pro

Security checks across malware telemetry and agentic risk

Overview

This spreadsheet skill is purpose-aligned overall, but it needs review because it runs local tooling and permanently modifies the user's LibreOffice macro profile without adequate scoping or cleanup.

Install only if you are comfortable with the agent running local Python, LibreOffice, and spreadsheet validation commands. Use a sandbox or disposable LibreOffice profile, back up existing LibreOffice macros before recalculation, and verify the missing MiniMaxXlsx executable before allowing the skill to invoke it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill instructs the agent to use file read/write and shell-executed tooling (`ipython`, `recalc.py`, LibreOffice headless, and the `MiniMaxXlsx` CLI) but does not declare permissions or narrowly scope when those capabilities may be used. This creates a capability/permission mismatch that can lead to unreviewed filesystem access and command execution during ordinary spreadsheet tasks, increasing the blast radius if the skill is invoked on untrusted inputs or modified tooling.

Vague Triggers

High
Confidence
97% confidence
Finding
The description says to engage for 'any task that involves tabular data, numeric analysis, or spreadsheet generation,' which is an overly broad trigger surface. Broad auto-activation can cause the skill to run in contexts the user did not intend, unnecessarily exposing shell and file capabilities and increasing the chance that adversarial content is processed through powerful local tooling.

Natural-Language Policy Violations

Medium
Confidence
80% confidence
Finding
The skill states that certain keyword signals automatically activate PivotTable behavior, including multilingual trigger terms, without requiring user confirmation. Forcing behavior from heuristic language triggers can override user intent and push the agent into using additional tooling and irreversible workflow constraints, such as the CLI pivot step that must be final.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The script silently writes a LibreOffice Basic macro into the user's profile-level macro directory, permanently modifying the user's office environment without consent or isolation. In a skill designed to process spreadsheets, this is more dangerous because it establishes persistent executable content in a trusted application context, which can surprise users, bypass expectations, and expand the blast radius beyond the current file being processed.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.