ClawHub
Back to skill

Security audit

BigFood · 冰箱管家（食材识别·智能食谱·菜谱·智能菜谱）| Fridge to Table & Recipe AI

Security checks for vulnerabilities and agentic risk

Overview

This is a recipe and fridge-photo analysis skill with no executable code, external backend, credential use, or cross-session persistence.

Installers should be aware that this skill may trigger on broad Chinese food or cooking phrases. It should be used for recipe help with user-provided food photos or ingredients; do not rely on its freshness or calorie estimates as precise health or safety advice.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • 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 (6)

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: bigfood
description: "🍽️ BigFood · AI冰箱管家 — 拍食材图片识别 → 智能推荐菜谱。支持多图混搭、冰箱食材管理、采购提醒、做饭灵感。食材识别、菜谱推荐、家常菜做法、冰箱库存管理一站式。| Fridge to Table — food recognition, recipe recommendation, ingredient & meal planning."
triggers:
  - 冰箱
  - 吃什么
  - 菜谱
  - 做饭
  - bigfood
  - 冰箱管家
  - 有啥菜
  - 食材
  - 今晚吃啥
---

# 🍳 BigFood · 冰箱管家

用户发食材图片，你直接用 vision 能力识别食材，然后推荐菜谱。不需�

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger list contains broad everyday phrases such as '冰箱', '吃什么', '菜谱', and '做饭' that can match many ordinary conversations unrelated to this skill. This can cause unintended invocation and context hijacking, where the cooking skill activates in situations the user did not clearly intend, increasing the chance of incorrect or intrusive behavior.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
96% confidence
Finding

The trigger '冰箱' is a single common noun and is too broad to safely gate skill activation. Any casual mention of a fridge could invoke the skill, creating accidental routing and unnecessary access to user context or images.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
96% confidence
Finding

The trigger '菜谱' is overly generic and commonly used in many harmless contexts. Because it lacks specificity, it may activate the skill for broad cooking discussions where the user did not intend to use this particular fridge-analysis workflow.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
96% confidence
Finding

The trigger '做饭' is a very common phrase in ordinary conversation and does not uniquely identify this skill. This can lead to accidental invocation during general meal planning or lifestyle discussion, reducing trust and increasing unintended behavior.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
96% confidence
Finding

The trigger '食材' is too short and generic, making it likely to match unrelated food conversations. This broad matching can cause the skill to activate without clear user intent and interfere with normal assistant behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.