Grocery

Security checks across malware telemetry and agentic risk

Overview

This grocery helper is purpose-aligned and transparent about storing local grocery, pantry, and household preference files, with privacy considerations users should understand.

Install only if you are comfortable with the assistant keeping grocery-related personal data in ~/grocery/. Review or delete those markdown files if you no longer want stored dietary restrictions, household notes, pantry inventory, or shopping history retained.

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)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill explicitly persists sensitive household context such as dietary restrictions, pantry contents, shopping history, and store preferences under ~/grocery/ without any user-facing notice about retention, sensitivity, or local file storage. While this is not an exploit primitive by itself, it creates a privacy risk because users may disclose health-adjacent and behavioral data that remains on disk across sessions and could be exposed to other local users, backups, or later tooling.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The skill exposes very broad natural-language examples like "Add milk" and similar everyday phrases without any explicit invocation boundary, which can cause accidental triggering during normal conversation. In a grocery assistant this is less severe than in a financial or system-control skill, but it can still modify user data unexpectedly, corrupt lists, and create confusion or unwanted purchases.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The documented behavior allows destructive actions like "Clear my list" and "Start fresh" to immediately archive and empty the current list without a confirmation step. Because these phrases are simple and plausible in ordinary conversation, an unintended invocation could erase active shopping data and disrupt household planning, even if recovery from history may be possible.

Missing User Warnings

Low

Confidence: 89% confidence
Finding: The template instructs the agent to create and populate local files containing household composition, dietary restrictions, picky-eater details, and shopping history without any privacy notice, consent prompt, retention guidance, or protection recommendations. While this is not an exploit by itself, it can lead to unintended storage of sensitive household and health-adjacent data on disk where other local users, backups, or synced services may access it.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal