Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording, but I found no executable code, persistence, credential handling, or hidden behavior.
Installers should treat this as a low-impact workflow helper, but should narrow or disable the generic trigger terms before broad deployment so it does not activate during unrelated security, GitHub, or bug-fix conversations.
The skill description uses very broad trigger terms such as "security" and "first," which are common in unrelated user requests. This can cause unintended invocation of the skill in contexts where it was not requested, potentially injecting workflow guidance or authority into unrelated tasks and degrading routing safety.
The keyword list contains ambiguous single words and generic context terms, including "security," "first," "before," and "github." These are likely to match a large volume of benign conversations, causing over-triggering and increasing the chance that the skill is invoked inappropriately or manipulatively.
The trigger list includes very common terms such as 'security', 'first', 'before', 'github', and 'bug fix', which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that this skill overrides a more appropriate skill or injects workflow guidance into contexts where it was not intended, reducing reliability and potentially affecting security-sensitive routing decisions.
The trigger sentences include broad, everyday terms such as "security," "first," "before," and generic help phrasing that can cause the skill to activate in contexts far beyond its intended purpose. Over-broad activation can route unrelated user requests into this workflow, creating confused-deputy behavior, irrelevant security guidance, or unintended handling of sensitive tasks under the wrong skill context.
The trigger examples are broad, fragmentary, and overlap with generic terms like 'security', 'first', 'github', and 'bug fix', which can cause the skill to activate in contexts the user did not intend. In an agent ecosystem, ambiguous activation increases the chance of misrouting requests, unnecessary invocation on sensitive tasks, or social-engineering-style prompt capture by making the skill claim relevance too often.
The trigger phrases are broad and generic, including terms like "security," "github," and "bug fix," which can cause the skill to activate in contexts far beyond its intended scope. In an agent ecosystem, overbroad activation can misroute unrelated user requests into this workflow, creating prompt-surface expansion, incorrect handling of sensitive tasks, and higher risk of unsafe or confusing automation.
The example trigger sentences are vague and do not establish clear activation boundaries, so they reinforce broad matching rather than constraining it. Because they resemble fragments that could appear in many normal requests, they make accidental or excessive invocation more likely.
文件名和正文内容表明该技能为中文（zh-CN）版本，但文中未看到向用户提供语言选择、回退机制或明确的语言适用说明。根据策略，若技能默认强制特定语言而未提供用户选择，可能构成语言/区域策略问题。
The description says to use the skill whenever users ask for broad categories like work-productivity, security, or practical workflows, but it does not define clear boundaries for when this specific skill should be selected. That ambiguity can lead to unintended invocation and misrouting, especially in mixed environments with many overlapping skills.
The default prompt embeds a very broad invocation phrase and generic terms like 'help me' and 'work-productivity', which can overlap with normal user language and cause accidental or implicit activation. Because the policy also enables implicit invocation, this increases the chance the skill is triggered outside clear user intent, potentially injecting workflow behavior or security-vetting guidance into unrelated conversations.
The file labels this document and companion files as Chinese ('zh-CN') and English variants, but this README itself mixes languages without stating whether the user can choose their preferred language or locale. That can conflict with language-choice policy when a specific locale is implied without opt-in or justification.
