Security audit

Countdown Timer - 倒计时助手

Security checks for vulnerabilities and agentic risk

Overview

This is a simple Chinese countdown/reminder helper that stores local countdown data, but its reminder features appear overstated or unfinished.

Install only if you are comfortable with a Chinese-language helper saving countdown titles and dates locally. Do not rely on it for real reminders, alarms, or safety-critical tasks unless the publisher adds a working scheduler and notification path.

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
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
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 (9)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The README exposes the skill through very broad natural-language examples such as reminders, countdowns, and pomodoro requests without defining clear invocation boundaries, exclusions, or disambiguation behavior. This can cause over-triggering or unintended interception of ordinary user requests, leading to incorrect actions such as creating reminders or storing dates when the user did not explicitly intend to invoke this skill.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The trigger keywords are broad everyday phrases such as asking how many days remain or mentioning timers, which can cause the skill to activate in contexts where the user did not clearly intend persistent storage or reminder behavior. In an agent system, overbroad activation increases the chance of unintended state changes, confusion, or privacy-impacting actions like saving dates or scheduling reminders without sufficiently explicit consent.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The scenario descriptions say when the skill might be useful but do not define firm activation boundaries or exclusion cases. That ambiguity makes accidental invocation more likely, especially because the skill can write files and register timed reminders, so a benign user question could lead to persistent actions the user did not expect.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill declares file read/write and scheduled reminder capabilities, but the user-facing description does not clearly warn that personal dates will be persisted or that future notifications may be generated automatically. This creates a transparency and consent problem: users may disclose birthdays or personal events without understanding they will be stored on disk and used for later reminder behavior.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: The file’s user-facing description and interaction strings are all Chinese-only, and there is no indication that the user can select another language or that the skill is intentionally limited to a Chinese-speaking context. This is a natural-language policy concern because it forces a locale without opt-in.

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The skill tells users '25 分钟后我会提醒你休息' but the implementation only returns a static text response and contains no scheduler, background task, or notification mechanism. This is a capability-deception issue: users may rely on a reminder that will never occur, causing missed tasks or unsafe assumptions about time-sensitive workflows.

Natural-Language Policy Violations

Low

Confidence: 88% confidence
Finding: All functional descriptions and usage examples are written in Chinese, and there is no indication that other languages are supported or that the language restriction is intentional for a region-specific use case. This can violate language/locale policy when a skill forces one language without explicit user opt-in.

Natural-Language Policy Violations

Low

Confidence: 84% confidence
Finding: 整个技能标题、触发规则、功能描述和示例对话均仅以中文呈现，未说明这是面向特定中文区域用户的限定技能，也未提供用户可选择其他语言的机制。若组织要求避免未经用户同意强制特定语言，这属于自然语言层面的策略风险。

Intent-Code Divergence

Low

Confidence: 88% confidence
Finding: The help text advertises countdown calculations for natural-language queries like '距离春节还有多少天', but the code does not implement holiday/date-name parsing or direct date-difference query handling. This can mislead users about the skill's capabilities and produce incorrect reliance on unavailable functionality, though the security impact is limited.

Static analysis

No suspicious patterns detected.