Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-style workflow helper with overly broad activation wording, but it does not request execution, credentials, persistence, or hidden access.
Before installing, be aware that this skill may activate on generic PDF or editing requests because its triggers are broad and implicit invocation is enabled. It appears safe as a workflow helper, but users or maintainers should narrow activation criteria if they want predictable routing.
The description says to use the skill when a user asks for broad categories like 'work-productivity', 'pdf', 'edit', or a generic 'practical workflow, artifact, checklist, analysis, or implementation support'. That invocation condition is so expansive that many unrelated tasks could match, leading to unintended activation and unsafe overreach of this skill's scope.
The trigger list includes extremely common terms such as 'nano', 'pdf', 'edit', 'natural', and 'language', which creates a high risk of accidental or overly frequent invocation. In an agent environment, broad matching can route unrelated user requests into this skill, causing misapplication of instructions, reduced reliability, and possible prompt-surface expansion into contexts the skill was not intended to handle.
触发关键词包含非常常见且语义宽泛的词，如“nano”“pdf”“edit”“natural”“language”“instructions”“bug fix”，会让该技能在与其真实用途无关的普通对话中被误触发。误触发会把用户请求错误路由到该技能，造成越权式上下文介入、错误建议或对其他更合适技能的遮蔽，尤其在多技能自动选择环境中更危险。
The default prompt is phrased so broadly that it can match ordinary user language about productivity, PDFs, editing, workflows, checklists, or implementation help, causing the skill to activate when the user did not explicitly request it. In combination with agent orchestration, this can silently inject the skill’s instructions and alter responses, creating prompt-scope confusion, unintended data exposure to the skill context, and unreliable behavior.
Enabling implicit invocation without strict activation constraints allows the platform to call this skill based on ambiguous conversational cues rather than clear user consent. Because this skill advertises a wide set of common tasks, implicit activation materially increases the chance of accidental invocation, instruction injection into unrelated conversations, and over-broad access to user context.
The trigger sentence is broad and conversational enough that normal user requests could accidentally match it, causing unintended skill invocation. In an agent ecosystem, ambiguous activation can route user inputs into the wrong workflow, leading to confusing behavior, incorrect document handling, or execution of steps the user did not explicitly request.
This trigger uses a generic request pattern without clear boundaries, so many unrelated requests could be interpreted as calls to this skill. Because the skill targets PDF and workflow assistance, accidental invocation could expose users to unintended processing paths, wasted agent actions, or mishandling of sensitive work-product context.
The trigger phrases are excessively broad and include common terms such as 'pdf', 'edit', 'natural', and 'language', plus generic starter phrases like 'Help me' and 'I need a practical workflow'. This can cause the skill to activate for unrelated everyday requests, leading to unintended invocation, prompt/context hijacking opportunities, and unsafe cross-task interference in agent workflows.
The example triggers are broad, repetitive, and lack any exclusion criteria, reinforcing permissive matching behavior. While not directly malicious, these examples train operators or routing systems to activate the skill on vague paraphrases instead of well-scoped intents, increasing accidental invocation risk.
该技能文件整体为 zh-CN 中文版本，内容未见向用户提供语言/区域选择，也未说明仅适用于中文用户或特定区域场景。按组织语言/locale 政策，强制单一语言而无用户选择可能构成自然语言策略违规。
技能描述中的启用条件使用了较开放的表述，例如“当用户提出 work-productivity, nano-pdf, nano, pdf, edit，或需要围绕该需求获得实用流程……时使用”，边界不清晰，容易被解释得过宽。这样的模糊匹配会放大自动选择偏差，使技能在不相关请求中被激活，降低系统可预测性并增加错误操作风险。
The trigger sentence is excessively broad and matches common, natural user phrasing such as 'Help me' or 'I need a practical workflow,' which can cause the skill to activate unintentionally outside its intended scope. In an agent environment, overbroad activation can route unrelated PDF or productivity requests into this skill, causing incorrect tool selection, confusing outputs, and increasing the chance that downstream actions are taken in the wrong context.
This file is presented as a Chinese README, but substantial user-facing descriptive content remains in English, which can impose a language choice on users without opt-in. The document does not indicate that users may choose their preferred language or that the mixed-language presentation is intentional.
