Security audit

智能随机：从8个维度、28个话题中均匀随机选择 - ⏰ 定时触发：可配置的心跳机制，实现定时随机提问 - 🎨 多维覆盖：情绪、身体、思维、行动、关系、环境、反思、未来 - 🔧 高度可定制：支持话题库扩展、触发概率调整、个性化设置 - 📊 使用统计：详细的统计报告和图表分析 - 🔄 上下文感知：结合对话历史和个人档案

Security checks for vulnerabilities and agentic risk

Overview

This is a self-reflection question skill with disclosed periodic prompts and personalization, but users should review its context and retention settings before enabling it.

Install this only if you want periodic personal reflection prompts. Before enabling it, review or change the trigger probability, active hours, personal-context use, previous-answer memory, 90-day retention, and backups. Do not run referenced install scripts or helper scripts unless you have the complete package and trust those files.

SkillSpector

Vulnerability Patterns

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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The README explicitly claims the skill uses conversation history and personal profiles for context awareness, but it does not disclose what data is accessed, how it is stored, or how users can disable it. This creates a real privacy risk because users may enable or install the skill without informed consent regarding potentially sensitive personal data processing.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The manual trigger phrases are broad natural-language utterances such as '随机问我一个问题' and '我想被提问一下', which could plausibly appear in ordinary conversation and accidentally invoke the skill. In an agent environment, ambiguous activation can cause unintended prompting, context use, or logging behavior without a sufficiently explicit user command.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The skill advertises timed and event-based triggering but does not define clear boundaries, conditions, or safeguards for when autonomous activation occurs. That ambiguity increases the risk of unsolicited prompts, repeated interruptions, and invocation in contexts where the user did not expect the skill to run.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The configuration enables use of personal context and retention of previous answers, but the skill description does not clearly warn users that potentially sensitive reflective data may be stored or reused. Because the prompts solicit emotional state, habits, relationships, and future intentions, silent retention can create meaningful privacy and profiling risk.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: This skill is configured to auto-load and run on a periodic basis with a 30-minute minimum interval and a 30% trigger probability, while also enabling use of personal context. That creates a broad, persistent activation surface that can cause unsolicited processing of user data and repeated interaction without clear per-context consent or exclusion boundaries.

Static analysis

No suspicious patterns detected.