Security audit

Yq Career Future Mirror

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent career-planning assistant, but it silently saves sensitive dialogue and creates personal HTML reports that load third-party resources.

Review before installing. Use this only if you are comfortable sharing resume-like details, income expectations, and personal career doubts with the assistant. Ask it not to silently save conversation logs, delete output/data/conversation_log.json after use, and be aware that generated HTML reports may contact third-party CDN and font services when opened online.

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 (4)

Vague Triggers

Medium

Confidence: 84% confidence
Finding: 技能元数据中的触发关键词覆盖“职业规划、职业转型、职业困惑、职业蓝图”等宽泛表达，容易与普通咨询类对话重叠，导致在用户并未明确请求该技能时被意外激活。这会把用户带入高数据收集流程，并触发后续文件写入、搜索和长期对话记录等行为，扩大隐私与越权操作风险。

Vague Triggers

Medium

Confidence: 87% confidence
Finding: 阶段一仅以“用户表达职业规划/转型/困惑的意向”作为触发条件，表述模糊且主观，给代理留下过大解释空间。这样的设计容易让模型在低置信度场景下启动敏感问卷收集，进而采集教育、工作经历、收入预期等个人信息。

Missing User Warnings

High

Confidence: 97% confidence
Finding: 文档明确要求在对话结束后“静默地”将用户与“未来自我”的问答追加到 conversation_log.json，但没有任何告知、同意或保留期限说明。这属于未披露的持久化个人数据处理，尤其该技能会收集职业经历、收入预期、困惑等高度敏感的个人背景信息，存在明显隐私与合规风险。

Ssd 3

Medium

Confidence: 95% confidence
Finding: 该技能要求把自然语言对话历史持久写入 JSON 文件，内容可能包含简历、工作经历、焦虑、收入预期和职业决策等敏感信息。由于没有最小化、脱敏、访问控制或生命周期管理要求，这种日志留存会放大二次泄露、误共享和越权访问风险。

Static analysis

No suspicious patterns detected.