Security audit

foolproof-kitchen

Security checks across malware telemetry and agentic risk

Overview

The skill is coherent as a meal-planning tool, but it needs Review because it handles sensitive health/body data, saves it into an HTML page, and loads third-party JavaScript without adequate disclosure or safeguards.

Install only if you are comfortable sharing body metrics and health-related dietary details with the agent and having the generated plan saved as a local HTML file. Review generated advice carefully, especially for pregnancy, diabetes, hypertension, gout, allergies, or other medical conditions, and prefer a version that bundles chart assets locally and adds clear privacy, retention, and medical-safety notices.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The template loads Chart.js from a third-party CDN, which introduces a supply-chain trust dependency into a page that displays sensitive user health and diet profile data. If the CDN, package, or delivery path is compromised, arbitrary JavaScript could execute in the user's browser and access all rendered personal data on the page.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The skill explicitly asks for sensitive health-related and body-metric data, including conditions such as diabetes, hypertension, pregnancy, and allergies, but it does not provide any user-facing notice about sensitivity, minimization, retention, or the limits of medical/nutritional guidance. In this context, the risk is elevated because users may disclose highly sensitive personal health information to obtain meal plans, and the skill presents itself as scientifically grounded, which may increase trust and disclosure.

Missing User Warnings

Low

Confidence: 83% confidence
Finding: The skill instructs the agent to save a generated HTML file into the workspace and present it to the user, but it does not warn the user that a local artifact will be created. Because the generated page may include sensitive body metrics, BMI, conditions, and nutrition targets, silently persisting the output can expose private information to other workspace users, logs, backups, or later sessions.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: This page handles body metrics and special-condition diet information, which can be sensitive health-related data, while also importing third-party JavaScript without any visible disclosure or consent flow. Even if the page itself does not intentionally exfiltrate data, the external script gains execution in the same context and could read or transmit the rendered data.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The file gives condition-specific dietary advice for diabetes, hypertension, gout, pregnancy, fat loss, and muscle gain without any visible warning that it is general educational guidance rather than medical advice. In this skill’s context, users are expected to provide personal body metrics and special dietary needs, so the content may be treated as individualized health guidance and could cause harm if followed without clinician review, especially for pregnancy or chronic disease management.

Missing User Warnings

Low

Confidence: 95% confidence
Finding: The recipe includes microwave heating guidance involving wrapped food and nearby egg-based preparations without any warning about steam release, superheating, or handling hot contents. In a consumer cooking skill aimed at beginners, omission of burn-safety instructions can plausibly lead to minor kitchen injuries, especially when users remove hot food immediately after microwaving.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.