Security audit
Security checks for vulnerabilities and agentic risk
This is a prompt-only workflow helper with overly broad activation wording, but no hidden execution, credential access, persistence, or destructive behavior was found.
This skill is reasonable to install if you want a general proactive-workflow helper, but expect it may be invoked too often because it allows implicit invocation and uses generic keywords like task, needs, transform, and bug fix. Review or narrow the triggers if precise routing matters in your environment.
The trigger phrases are broad, generic help-seeking language that can cause this skill to activate for many unrelated user requests. In an agent ecosystem, overbroad routing can misapply workflow instructions, override more appropriate skills, or cause unintended autonomous behavior in contexts the user did not explicitly request.
The manifest description instructs use for broad terms like 'work-productivity', 'proactive', 'transform', and 'task', which are common across many benign conversations. Ambiguous invocation guidance increases the chance that the skill is selected outside its intended scope, creating prompt-routing confusion and raising the risk of irrelevant, misleading, or lower-safety behavior being applied to unrelated user requests.
The trigger list is overly broad and includes generic words such as 'task', 'needs', and 'partners' that commonly appear in unrelated user requests. This can cause unintended skill activation, leading the agent to inject workflow-specific guidance into contexts where it was not requested, reducing reliability and potentially bypassing safer or more appropriate skill routing.
The manifest uses very broad activation wording in the display and prompt text, while also enabling implicit invocation. That combination can cause the skill to be selected in loosely related user requests, increasing the chance of unintended execution, prompt-scope confusion, or misuse in contexts the skill was not narrowly designed for.
The trigger sentences are broad, natural-language phrases that can match many ordinary user requests unrelated to this specific skill. That increases the chance of accidental invocation or misrouting, which can cause the agent to apply the wrong workflow, expose irrelevant capabilities, or bypass more appropriate domain-specific handling.
The trigger keywords and example activation phrases are very broad, including generic terms like 'task', 'transform', 'needs', and 'bug fix'. This can cause the skill to activate outside its intended scope, increasing the chance of accidental invocation, context hijacking, or interception of unrelated user workflows in an agent environment.
The file descriptions state that SKILL.md and README.md are the English versions while this file is the Chinese version, but there is no visible note offering the user a language choice or opt-in within this README. This creates a natural-language locale constraint that may conflict with a policy requiring user language selection rather than assuming a fixed locale.
该文件为 SKILL.zh-CN.md，全文内容固定为中文，但没有自然语言说明用户可选择其他语言或这是仅面向中文用户的受限版本。按规则，若技能对语言/locale 施加约束而未提供选择或正当说明，属于策略层面的自然语言问题。
技能启用条件描述覆盖了 "work-productivity, proactive-agent, proactive, transform, task" 以及“需要实用流程、产物、检查清单、分析或实现支持”等宽泛需求，边界非常模糊。这会导致技能在大量普通生产力或实现类请求中被激活，使系统更难预测何时调用该技能，并增加被不相关或对抗性输入利用的机会。
触发关键词包含 "proactive"、"task"、"needs" 等高频日常词汇，容易在与技能目标无关的普通对话中误触发。误触发会让代理在错误上下文中加载该技能，扩大提示注入面并造成不必要的行为偏移、混淆或越权辅助。
The keyword list includes very generic terms such as 'task', 'transform', 'partners', and 'needs', which are likely to appear in benign everyday prompts. In a routing or auto-selection system, this can over-activate the skill and create unsafe scope expansion, where the agent applies this capability in contexts the user did not intend.
The file lists separate English and Chinese instruction and README files, but does not state how the skill selects a language or whether the user can choose their preferred locale. This can create an implicit language/locale behavior without documented opt-in, which may conflict with organizational language-choice expectations.
示例触发句只展示了正向例子，而且内容还是对需求文案的截断复述，缺少边界案例与反例说明。结果是集成方或路由器更可能采用宽松匹配，进一步放大误触发问题。
