Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with broad activation wording but no hidden execution, credential access, persistence, or data exfiltration.
Before installing, be aware that this skill may be invoked too broadly for general PDF, editing, or language-related requests. Prefer explicit invocation by skill name when you want this workflow helper, and review outputs for relevance if it activates implicitly.
The trigger sentence is broad, natural-language, and closely resembles an ordinary user request rather than a narrowly scoped invocation pattern. This can cause unintended skill activation in unrelated conversations, leading the agent to apply this workflow when the user did not explicitly request it, which is a prompt-routing and control-boundary weakness.
The manifest description says to use the skill when a user asks for broad terms such as 'pdf', 'edit', or a 'practical workflow, artifact, checklist, analysis, or implementation support,' which are common across many benign tasks. Such ambiguous activation criteria increase the chance of unintended invocation and allow this skill to intercept requests far beyond the narrow Nano Pdf workflow context.
The trigger keyword list is excessively broad and includes generic terms like 'nano', 'pdf', 'edit', 'natural', and 'language' that commonly appear in unrelated user requests. This can cause the skill to activate outside its intended scope, leading to prompt hijacking of normal conversations, misrouting, and accidental application of this workflow where it does not belong.
The front-matter description says to use the skill whenever users mention broad terms or need almost any practical workflow, checklist, analysis, or implementation support around the requirement. This ambiguous scope can make the skill match far beyond its intended domain, causing accidental invocation and reducing routing integrity across the agent system.
The trigger list includes very common words such as "nano", "pdf", "edit", "natural", and "language", which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that users are routed into an unintended workflow, producing irrelevant guidance or interfering with safer, more appropriate skills.
The trigger sentences are broad, natural-language phrases that can match ordinary user requests unrelated to this specific skill, causing over-activation or misrouting. In an agent environment, this can lead to the skill being invoked when the user did not intend it, increasing the chance of inappropriate context capture, workflow interference, or unsafe downstream actions if the skill later performs edits or automation.
The invocation guidance mixes generic keywords and ambiguous natural-language examples without defining strict trigger conditions. In an agent ecosystem, this increases the chance of accidental or over-broad routing, which may override user intent, invoke the wrong workflow, or expose downstream tools/processes unnecessarily.
The trigger phrases are broad enough to match ordinary requests about PDFs, editing, or generic workflow help, which can cause the skill to activate outside its intended scope. In an agent ecosystem, this increases the chance of unintended delegation, prompt hijacking of unrelated tasks, or misuse of the skill in contexts where its assumptions and safeguards do not apply.
The example trigger sentences model vague invocation patterns and repeat broad natural-language phrasing without clear scope checks. In practice, examples like these train routing systems or authors toward permissive matching, which amplifies accidental activation and increases the surface for skill misuse or conversation hijacking.
The file is a Chinese-localized skill document, but core descriptive content and user-facing requirement text remain in English. This can impose a mixed-language experience without explicit user opt-in or a documented language-selection mechanism.
The default prompt contains broad, common-language trigger terms and directly encourages invocation for generic productivity/PDF-related requests. Combined with allow_implicit_invocation=true, this can cause the skill to activate in situations the user did not clearly intend, increasing the risk of prompt-scope overreach, confused-deputy behavior, or unnecessary exposure of the skill's instructions and capabilities.
A file labeled as Simplified Chinese contains key descriptive and trigger content in English, including the main workflow description and activation phrases. This can effectively impose an unexpected language mode on users without opt-in or justification, which is a natural-language locale policy concern.
