ClawHub
Back to skill

Security audit

Recipes

Security checks across malware telemetry and agentic risk

Overview

This looks like a recipe lookup skill, but it asks users to install and run executable code from an external GitHub repo that is not included in the reviewed bundle.

Treat this as a Review item, not known malware. The recipe lookup purpose is coherent, but do not run the README install commands blindly because the executable CLI code is outside the reviewed package and is pulled from a mutable GitHub repo. Inspect the referenced repository and scripts first, avoid running them with elevated privileges, and prefer a package that includes or pins the executable code being installed.

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 (1)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill's activation guidance says 'When user asks about recipes/cooking,' which is broader than the concrete command set and may cause the agent to invoke this skill for loosely related food queries. Overly broad trigger language can lead to unnecessary tool use, context leakage into external API requests, or reduced reliability if the agent routes non-recipe cooking questions to a skill that cannot answer them well.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.