ClawHub
Back to skill

Security audit

Lab Budget Forecaster

Security checks across malware telemetry and agentic risk

Overview

This appears to be a simple local lab budget forecasting skill, with some broad wording but no evidence of hidden access, network use, persistence, or destructive behavior.

Reasonable to install for lab budget forecasting. Use it only with intended budget inputs and expense CSVs, and check the script help because the documentation and actual command-line arguments differ.

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 (4)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The skill description is overly broad for a tool that ultimately executes a local Python script, so it may be auto-selected for generic data-analysis requests outside its narrowly intended budgeting use case. That increases the chance of unintended code execution, misuse on inappropriate inputs, and confusion about the skill's true safety boundaries.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The 'When to Use' section repeats generic language about structured data analysis and bounded outputs rather than clearly constraining the skill to laboratory budget forecasting. In an agent environment, ambiguous trigger guidance can cause the skill to be invoked for unrelated tasks, exposing local script execution and file I/O capabilities in contexts the author did not intend.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The invocation label is broad enough to match many generic data-analysis requests, which can cause the agent to activate this skill outside its intended budget-forecasting niche. That increases the chance of unintended tool routing, scope drift, and outputs presented with inappropriate authority, especially since the same audit already records a stress-case boundary failure.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The alternate invocation text is similarly underspecified and promotes activation for generic tasks needing assumptions and reproducible output, which describes a very large class of requests. In context, this is more concerning because the skill is packaged as 'Production Ready' despite documented scope-boundary weakness in stress scenarios, so overbroad triggering can expose known failure modes more often.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.