Back to skill

Security audit

nutrition-and-health

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent nutrition guidance skill with no observed hidden execution or network behavior, but users should be careful about sharing personal health details.

Install only if you want a Chinese-first nutrition and wellness assistant. Avoid sharing medical details you would not want stored or logged by the surrounding agent platform, and treat condition-specific advice as general wellness guidance rather than medical diagnosis or treatment.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (18)

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The README advertises very broad trigger coverage for generic health and food phrases, which can cause the skill to activate in many ordinary conversations where the user did not intend to invoke a health advisor. In a health-related context, accidental invocation is more dangerous because it can collect sensitive profile information or provide quasi-medical guidance without clear user intent.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 366)May include surrounding context.

md
- `SKILL.md` - 本文件,使用说明

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README states '中文优先' and the closing tagline says the skill is made for Chinese users, but there is no indication that users can choose another language or locale. This is a natural-language policy concern because it imposes a language preference without documented opt-in or a clearly justified hard regional constraint.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The README states that the skill can 'build a personal profile' and uses local storage, but it does not clearly disclose what health-related data is collected, how long it is retained, or how the user can review and delete it. Because the skill targets nutrition, wellness, and potentially chronic-condition scenarios, the stored data may be sensitive and create privacy risk even if it is not uploaded remotely.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The description and most user-facing instructions are written as Chinese-only guidance, while the file does not state that users may choose their preferred language. Although an English trigger section exists, the skill presentation still defaults to Chinese without explicit language selection, which can violate a language-choice policy.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases in this section are very broad everyday expressions, so the skill may activate during normal conversation even when the user did not intend to invoke a health-advice workflow. In a nutrition/health context, unintended activation is more concerning because the skill may begin collecting sensitive health profile data or provide quasi-medical guidance without clear user opt-in.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The English activation examples are also common conversational phrases, which increases the chance of accidental invocation in unrelated chats. Because this skill covers health, diet, and body goals, accidental activation can expose users to unsolicited health recommendations or initiate collection of sensitive personal attributes.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

The manifest description markets the skill as a general-purpose personal health advisor covering many everyday topics such as diet, beverages, fitness, and daily health care. Because it does not define specific trigger phrases, boundaries, or exclusion cases, it risks matching common conversational requests too broadly.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

This JSON file stores all user-facing drink names, descriptions, and situations exclusively in Chinese, with no indication that language selection is optional or region-specific. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The script solicits health status, allergies, age, sex, height, weight, habits, and goals, which together constitute sensitive personal and health-related data. Although it says users may skip questions, it provides no privacy notice, purpose limitation, retention/deletion policy, or handling guidance, creating a real privacy risk if this data is logged, stored, or reused by the surrounding skill framework.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger guidance uses very generic everyday phrases like '今天吃什么' and '早餐吃什么', which can cause unintended invocation in ordinary conversation. In a health/nutrition skill, accidental activation is mainly a privacy and UX concern because users may unexpectedly engage a health-oriented workflow or reveal personal preferences and conditions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

Examples such as '我胃不好' and related recommendation prompts invite health-related guidance without clearly bounding the skill to general wellness rather than condition-specific advice. In this context, ambiguous scope can lead users to over-rely on the skill for symptoms or medical issues, increasing the risk of unsafe or inappropriate recommendations.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The documentation makes a strong privacy and data-handling guarantee ('all information is stored locally' and 'never uploaded') without any verifiable technical basis in this file. Users may disclose sensitive health information in reliance on that promise, creating privacy, compliance, and trust risk if the broader system transmits, logs, or syncs data elsewhere.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
67% confidence
Finding

The natural-language description is written only in Chinese, while the manifest lists both zh-CN and en as supported languages. This creates a potential locale-policy concern if the skill behavior or presentation defaults to Chinese without making language choice explicit to users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

This JSON file contains user-facing meal names, descriptions, and benefits predominantly in Chinese, which can impose a specific language on downstream users if surfaced directly. There is no accompanying indication in this file that the language is optional, selectable, or intentionally limited to a Chinese-speaking context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

This JSON file contains user-facing natural-language content exclusively in Chinese, but does not document that the skill is Chinese-only or offer any language/locale choice. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

All prompts, messages, and docstrings are written in Chinese, so the skill implicitly forces a specific language for interaction. The file does not offer opt-in language selection or explain that the skill is intentionally limited to a Chinese-speaking context, which conflicts with the language/locale policy criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

All user-facing instructions, examples, and interaction patterns are presented only in Chinese, with no indication that the user may choose another language. Under the policy, forcing a specific language without opt-in can be a natural-language policy violation unless the locale constraint is documented and justified.

Content

No source excerpt is available for this finding.