Back to skill

Security audit

今天吃什么 - 智能三餐规划引擎

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Chinese meal-planning skill that runs a local planner and keeps a small meal-history file for deduplication.

Install this if you want Chinese-language household meal planning and are comfortable with a local history.json file recording recent meal recommendations for deduplication. Use --no-save for testing without writing history, and use --reset-history only when you intentionally want to clear that local meal history.

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 Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding

The skill instructs the agent to run a local script and explicitly documents persistent history writes and reset behavior, but it does not declare any tool scope or file access boundaries. This creates an authorization gap where the agent may use file-write capable execution without transparent permissioning, increasing the risk of unintended file modification or abuse if the script or surrounding files are altered.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation phrases are extremely broad everyday requests such as '今天吃什么' and '不知道吃啥', which can cause the skill to trigger in many casual contexts without the user clearly consenting to script execution or persistence. In this skill, auto-triggering is more concerning because activation leads to program execution and possible history updates across sessions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill advertises cross-session deduplication via persistent history and exposes a reset-history operation, but provides no clear user warning about data retention, what is stored, how long it is kept, or the consequences of deletion. This is a privacy and transparency issue, especially because the stored history reflects recurring household behavior and can be modified or erased by the skill workflow.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown skill file presents all instructions and content in a single language, which can amount to a language/locale policy violation when no user opt-in or alternative language option is provided. The file does not indicate that it is intentionally limited to Chinese users or a region-specific context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module docstring, usage text, and overall skill framing are written entirely in Chinese and present the skill as producing Chinese meal-planning output, with no indication that the user can choose another language or locale. This is a natural-language locale constraint that can violate organizational language-choice policy when not explicitly optional or justified as region-specific.

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 contains natural-language fields and values specific to Chinese locale, such as weekday_cn and the festival name 端午节, indicating the skill data is tailored to a single language/locale. Under the policy, forcing a specific language or locale without user opt-in or documented justification 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
83% confidence
Finding

The entire reference file and its prescribed example copy are written only in Chinese, and it instructs the LLM to use Chinese holiday/solar-term phrasing in generated text. There is no indication that users can choose another language or locale, which may violate language/locale choice policy.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

This JSON dataset uses Chinese dish names, ingredient labels, flavor descriptors, and notes throughout, with no adjacent metadata indicating that Chinese is optional or that users can choose another language. Under the language/locale policy criterion, a file that effectively enforces one language without user opt-in can be considered a natural-language policy issue.

Content

No source excerpt is available for this finding.