Security audit
Security checks for vulnerabilities and agentic risk
This skill is a documentation/workflow helper with broad activation wording, but it does not include code, credential handling, persistence, or hidden high-impact actions.
Before installing, be aware that this skill may activate on ordinary Google, Workspace, CLI, Gmail, Calendar, Drive, Contacts, or bug-fix requests. It appears safe as a workflow helper, but a publisher should narrow the triggers or require explicit invocation to reduce accidental use.
The skill description is extremely broad and matches common terms like "google," "workspace," and "cli," which can cause the skill to activate for many unrelated requests. This creates routing confusion and increases the chance that an over-privileged or irrelevant workflow is invoked in contexts the user did not intend, which is a real security and safety boundary issue for agent systems.
The trigger keywords are overly generic and include highly common terms such as "google," "workspace," "cli," "gmail," "calendar," and "drive," without scope constraints. In an agent environment, this can cause accidental invocation on unrelated user requests involving common productivity tools, leading to inappropriate prompt injection of this skill's workflow and potential misuse of connected capabilities.
技能描述中的适用条件覆盖面过大，如“当用户提出 work-productivity, gog, google, workspace, cli，或需要围绕该需求获得实用流程、产物、检查清单、分析或实现支持时使用”，几乎可匹配大量普通生产力或开发请求。边界不清会导致技能选择器过度偏向该技能，增加错误调用和不相关响应的概率。
触发关键词包含非常宽泛的常见词，如“google”“workspace”“cli”“bug fix”，会让该技能在大量无关请求中被意外激活。误触发会扩大该技能对会话路由和响应内容的影响面，造成错误工具选择、用户意图偏离，甚至掩盖本应由更精确技能处理的任务。
The trigger sentence is written as a broad natural-language phrase that could match ordinary user requests instead of a narrowly scoped invocation. In an agent ecosystem, this can cause unintended skill activation, routing user data or actions into the wrong workflow and increasing the chance of confusing, unsafe, or unauthorized task execution.
The trigger sentence is broad enough to match ordinary user phrasing about getting help with workflows, bug fixes, or setup hardening, which can cause the skill to activate outside its intended scope. In an agent ecosystem, this increases the chance of unintended routing or prompt injection surface expansion because the skill may engage on loosely related requests involving Google Workspace or productivity tasks.
This activation pattern is similarly ambiguous and overlaps with common natural-language requests for practical workflows, making accidental invocation likely. Because the skill targets broad work-productivity and Google-related tasks, the surrounding context makes overmatching more dangerous by pulling in many normal requests that were not meant to invoke this specific skill.
The trigger phrases are broad enough to match common requests about productivity, Google Workspace, or bug fixing, which can cause the skill to activate outside its intended scope. In an agent ecosystem, overscoped routing can lead to the wrong workflow being invoked, producing unintended actions, misleading outputs, or unnecessary access to connected tools.
该技能文件以 SKILL.zh-CN.md 命名并全文使用简体中文（中国地区）表述，但内容中未说明这是可选语言版本，也未向用户提供语言/地区选择或切换方式。根据语言/地区策略，若无明确用户选择，固定单一语言/locale 可能构成自然语言政策问题。
The default prompt uses very broad terms such as work productivity, workflow help, fixing bugs, and hardening setup, which overlap with common user requests across many domains. In combination with implicit invocation, this creates a realistic risk of over-triggering the skill and applying its instructions in contexts where they were not intended.
The skill enables implicit invocation with no trigger constraints or exclusion logic, which can cause it to activate on unrelated user requests. Because the skill is framed around broad productivity and workflow help, this increases the chance of unintended routing, prompt injection exposure, or the skill influencing conversations outside its intended scope.
The invocation guidance is ambiguous because it mixes generic productivity terms with a wide requirement statement, without defining what requests should not trigger the skill. This makes accidental invocation more likely, especially in multi-skill environments where overlapping phrases can route tasks incorrectly and produce unintended access to user context or actions.
This file is presented as a Chinese README, but substantial user-facing content and trigger text remain in English, with no indication that the user can choose their preferred language. That can amount to an implicit language constraint or inconsistent locale behavior without opt-in.
No suspicious patterns detected.
