Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overbroad activation wording, but it does not request sensitive access, persistence, hidden execution, or external data handling.
Installers should know this skill may be selected too often because its trigger words include generic terms like task, needs, proactive, and bug fix. Review or narrow the trigger metadata if precise routing matters, but the inspected package does not show credential use, hidden execution, persistence, or destructive behavior.
The trigger phrases are extremely broad and match common productivity language such as 'help me', 'I need a practical workflow', and generic terms like 'task' or 'bug fix'. This can cause the skill to activate in many unrelated conversations, leading to unintended instruction takeover, user confusion, or inappropriate workflow injection in contexts where the skill was not requested.
The trigger phrases are broad enough to match common, non-specific requests such as asking for practical workflows, help, or task support. In an agent-routing system, this can cause the skill to activate outside its intended scope, leading to incorrect delegation, unintended processing of unrelated user requests, and expanded attack surface for prompt/skill hijacking through accidental invocation.
The skill description includes highly generic invocation terms such as 'task', 'analysis', 'workflow', and 'implementation support' that overlap with ordinary user requests. In auto-dispatch or retrieval systems, this can cause the skill to activate for unrelated conversations, expanding its influence beyond its intended scope and potentially steering users into inappropriate workflows.
The trigger list contains vague keywords like 'proactive', 'transform', 'task', 'needs', and 'bug fix', plus broad example phrases that resemble normal conversation. This materially increases the chance of unintended invocation, especially in systems that match on keywords, which can lead to prompt-routing errors, context hijacking, or inappropriate application of the skill in benign chats.
The trigger list includes very generic terms such as "task", "needs", and "bug fix", which are likely to appear in many unrelated user requests. This can cause unintended skill activation, leading the agent to apply this workflow in the wrong context and potentially override better-matched skills or produce irrelevant actions.
This localized README is labeled as Simplified Chinese, but the central behavior description and trigger phrases are presented only in English. That creates a language-policy concern because the skill appears to impose English interaction in a Chinese-language variant without documenting an opt-in or justification.
The invocation description says to use the skill for broad categories like work-productivity, proactive, transform, or task, without clear boundaries. This ambiguity increases the chance of accidental invocation for loosely related requests, which can reduce reliability and steer the agent into an inappropriate workflow.
The skill metadata uses broad, generic phrasing such as 'work-productivity', 'practical workflow', 'artifact', 'checklist', and 'implementation support', which can match many unrelated user requests. Combined with a vague default prompt, this increases the chance of unintended routing or activation, causing the skill to influence conversations outside its intended scope and potentially expose users to incorrect or unsafe assistance.
Enabling implicit invocation without tight trigger constraints allows the platform to auto-select this skill based on loosely related requests. In this skill, the broad productivity-oriented scope makes accidental invocation more likely, which can lead to prompt-space contamination, inappropriate workflow execution, or user confusion about why the skill was engaged.
The trigger section uses broad generic terms and sentence stems such as "task," "transform," and "I need a practical workflow," which can match many normal user requests outside the intended scope. In an agent routing system, this can cause the skill to activate unexpectedly, leading to inappropriate workflow injection, user confusion, or overshadowing of more suitable skills.
