Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with no executable payload, though its activation wording is broader than ideal.
This skill is reasonable to install if you want a Nano Pdf-style workflow helper. Be aware that its trigger words are broad, so it may appear for ordinary PDF, editing, or workflow requests where a narrower skill would be better.
The manifest description says to use the skill when a user asks for broad terms like "pdf," "edit," or a "practical workflow," which are common across many benign tasks and unrelated domains. This creates a high risk of accidental invocation and privilege overreach, because the skill may be selected for requests that have little to do with the narrowly intended Nano Pdf workflow use case.
The trigger list is overly broad and includes generic terms like "nano," "pdf," "edit," and even "natural" and "language," which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that this skill intercepts requests outside its intended scope, leading to misrouting, unintended disclosure of context to the skill, or unsafe/irrelevant guidance being surfaced.
The trigger keyword set is overly broad and includes generic terms like "nano", "pdf", "edit", "natural", and "language", which can match many unrelated user requests. This can cause the skill to activate outside its intended scope, leading to inappropriate routing, confused task handling, or accidental exposure of workflow behavior in contexts where it does not belong.
The trigger phrases are broad, generic help-request patterns that can overlap with ordinary user speech, increasing the chance that this skill is invoked unintentionally. In an agent ecosystem, accidental activation can cause the wrong workflow or instructions to run, leading to misrouting, confusing outputs, or unsafe application of PDF-related guidance outside the user's intent.
The trigger phrases are extremely broad, including generic terms like 'nano', 'pdf', 'edit', 'natural', and 'language', which can cause the skill to activate in contexts unrelated to the intended workflow. Overbroad activation boundaries increase the chance of unintended routing, prompt hijacking of unrelated user tasks, and misuse of the skill in contexts where its assumptions or outputs are unsafe.
The example trigger sentences reinforce vague activation by repeating broad requirement language without showing clear boundaries, disambiguation, or concrete user intents. In practice, this trains routing toward permissive matching and makes accidental activation more likely, though the examples themselves do not add new capabilities.
The skill description says it should be used when users ask for broad categories like work-productivity, nano, pdf, or need general workflow or implementation support, but it does not clearly define what is in or out of scope. This ambiguity increases the chance of unintended invocation and misclassification of user intent, which can degrade reliability and cause the wrong skill to handle a request.
The default prompt embeds a very broad natural-language invocation phrase (e.g. 'help me' plus common productivity terms) while implicit invocation is enabled, which increases the chance the skill is triggered by ordinary user speech rather than deliberate selection. This can cause unintended routing of user requests into the skill, expanding attack surface and making prompt-injection or unsafe workflow execution easier in unrelated conversations.
The trigger sentences are broad, natural-language phrases that can match ordinary user requests about PDFs, editing, or practical workflows, increasing the chance of unintended skill activation. In an agent environment, overbroad invocation can route tasks to this skill when the user did not explicitly intend it, causing incorrect task handling, context leakage across workflows, or bypass of more appropriate controls.
The file descriptions explicitly label the instruction and guide files as English and Chinese variants, but the README does not indicate that users can choose their preferred language or locale at runtime. This can conflict with a language/locale policy requiring user choice rather than an assumed default.
This file presents the user-facing description entirely in Chinese, but does not state that the locale is optional or limited to a Chinese-language audience. A user-facing skill artifact that effectively forces one language without opt-in can violate language/locale policy expectations.
The file includes non-English content such as Chinese link titles in the evidence list, but does not state whether multilingual output, language selection, or locale handling is user-driven. This can create an implicit language/locale constraint without opt-in, which falls under natural-language policy concerns.
