ClawHub
Back to skill

Security audit

Personal Nutrition

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent nutrition tracker, but it needs review because it can persistently read and update sensitive diet and weight notes from broad everyday triggers, and its README includes a typo-like external download link.

Install only if you are comfortable with the assistant keeping a local nutrition history that may include meals, calories, water intake, and weight. Use explicit logging language, review the stored nutrition file periodically, and verify the README's external download URL before following it.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • 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
Findings (4)

Direct Prompt Extraction

High
Category
System Prompt Leakage
Content
| Log weight | `вес сегодня 74.5кг` |
| Ask for summary | `что я ел сегодня?` |
| Ask for advice | `как улучшить завтрак?` |
| Show instructions | `инструкция` |

## Installation
Confidence
85% confidence
Finding
Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.

Vague Triggers

High
Confidence
96% confidence
Finding
The trigger phrases are extremely broad and include common words like food, drink, weight, and meal, which can cause the skill to activate in many ordinary conversations outside the user's intended nutrition-tracking context. Because the skill is instructed to read and update a persistent personal data file, over-triggering increases the chance of unwanted access to or modification of sensitive health-related notes.

Natural-Language Policy Violations

Medium
Confidence
87% confidence
Finding
The operational instructions and role description are written in Russian, which strongly implies the skill will interact in Russian by default. There is no indication that the user can choose another language or that the Russian-only behavior is justified by a documented locale constraint.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The skill directs the agent to update a persistent user file whenever food or water data is mentioned, but it does not require explicit user consent or even notify the user that durable storage will occur. In a personal-health context, this can silently accumulate sensitive dietary and weight information and can also cause incorrect records if casual conversation is mistaken for logging intent.

Static analysis

No suspicious patterns detected.