Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording, but it does not contain hidden code, persistence, credential handling, exfiltration, or destructive behavior.
Before installing, be aware that the skill may activate too broadly for generic productivity requests because of its trigger wording and implicit invocation setting. It appears safe as a low-impact workflow helper, but users who want precise routing should narrow the triggers or invoke it explicitly by name.
The manifest description defines activation using very broad request patterns such as 'work-productivity', 'proactive', 'transform', 'task', and general requests for practical support. This ambiguity can cause accidental routing of unrelated conversations into this skill, expanding its effective scope beyond what users likely intended.
The trigger list is excessively broad and includes generic terms like 'task', 'proactive', 'needs', and 'bug fix', which can cause the skill to activate for many unrelated user requests. Over-broad activation increases the chance that this skill is invoked out of context, leading to unintended behavior, user confusion, or interference with more appropriate skills.
The trigger list includes very broad everyday terms such as "task", "needs", and "partners", which can cause the skill to activate in many unrelated conversations. Overbroad activation increases the chance of unintended instruction injection into user workflows, confusing routing, and accidental disclosure or modification of context when the wrong skill is invoked.
The trigger examples are broad, natural-language phrases like 'Help me' and 'I need a practical workflow', which are common in ordinary user requests and can cause the skill to activate unintentionally. In an agent-routing context, this increases the chance of over-invocation, context confusion, and inappropriate execution of this skill when the user did not specifically intend it.
The trigger phrases are extremely broad and include generic terms like "task," "transform," and common help-request templates, which can cause the skill to activate in many unrelated conversations. In an agent setting, over-activation can misroute user requests, override more appropriate skills, and lead to unintended autonomous workflow behavior in contexts where the user did not explicitly request it.
This README presents the skill instructions and usage entirely in Chinese, including headings and trigger guidance, with no visible statement that users may choose another language. A language-specific constraint without opt-in can violate organizational language/locale policy for general-purpose skills.
The trigger keywords and example phrases are broad enough to match many generic productivity or task-oriented requests, which can cause unintended skill activation. In an agentic workflow context, accidental invocation can redirect user tasks, inject irrelevant workflow steps, or override a more appropriate skill, reducing reliability and potentially affecting safety-sensitive task routing.
文件名和正文整体为 zh-CN 中文区域版本，但文档中未说明这是可选语言版本，也未向用户提供语言选择或切换机制。按规则，若技能强制特定语言/地区而没有用户选择或明确合理说明，属于自然语言政策风险。
The description says to use the skill when a user asks for broad categories like work-productivity, proactive, transform, or task-related help, but it does not clearly define boundaries. This ambiguity can lead orchestration systems or operators to invoke the skill in situations it was not designed for, increasing the risk of inappropriate actions or misleading output.
The skill metadata and default prompt are highly generic and broad, matching common terms like work-productivity, proactive, transform, task, checklist, analysis, and implementation support. This increases the chance the skill is invoked in contexts beyond its intended scope, which can cause unintended prompt injection surface expansion, misrouting, or unsafe automation behavior.
Enabling implicit invocation without tight trigger constraints allows the platform to auto-select this skill for loosely related user requests. Combined with the broad metadata, this makes accidental or overbroad activation more likely, which can expose users to unintended actions, unreliable outputs, or abuse through prompt steering.
