Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only Word/DOCX formatting helper with overly broad activation wording, but no hidden code, credential use, persistence, or destructive behavior was found.
This skill is reasonable to install for Word/DOCX formatting cleanup help. Be aware that its automatic activation rules are loose, so users or administrators may want to narrow invocation to explicit Word/DOCX repair requests if many skills are installed.
The trigger sentences are so generic that they can match ordinary user requests about productivity, formatting, or workflows without clear intent to invoke this specific skill. Overly broad routing can cause the wrong skill to activate, increasing the chance of irrelevant guidance, prompt-space interference, or untrusted skill content being injected into otherwise normal conversations.
The trigger sentences are generic and map to broad productivity concepts like 'practical workflow' and 'help me', which can cause the skill to activate for loosely related Word or document requests without clear user intent. In an agent environment, overbroad activation increases the chance of the model pulling in this skill unnecessarily, leading to incorrect task routing, confusion, or execution of irrelevant guidance.
The trigger phrases are broad and generic enough to activate on common productivity or Microsoft Word requests that may not actually need this specific skill. Over-broad activation can cause inappropriate routing, unexpected tool use, or insertion of workflow behavior in unrelated contexts, which is a real security and safety concern for agent systems even though the file contains no direct exploit payload.
The trigger examples are written as broad natural-language requests that overlap with common productivity queries, which can cause the skill to activate when a user did not explicitly intend to invoke it. In an agent system with skill routing, this increases the chance of incorrect tool selection, prompt-scope expansion, and unintended handling of user content under this skill’s instructions.
The skill description says to use the skill whenever users ask for broad categories like work productivity, Microsoft Word, formatting, styles, or implementation support. Those activation boundaries are wide enough to match many ordinary requests, which can cause unintended invocation and route users into a skill they did not explicitly request. In an agent system, overbroad routing increases the chance of mis-scoped actions, confusing outputs, and unsafe delegation behavior.
The keyword list includes generic terms such as automation, styles, find replace, and document formatting, which commonly appear in everyday conversations outside the intended scope. Generic trigger terms create collision risk with unrelated tasks, making accidental activation more likely and reducing operator control over when this skill is used. In multi-skill environments, this can degrade routing integrity and expose users to unintended workflows.
The default prompt is extremely broad and effectively says to use the skill for a wide class of productivity and Word-related requests without any narrowing conditions. In combination with agent routing, this can cause the skill to be invoked in situations the user did not clearly intend, increasing the chance of inappropriate tool use, prompt-scope confusion, or unintended handling of sensitive document content.
Enabling implicit invocation without a narrowly constrained scope allows the system to route user requests to this skill automatically based on loose relevance signals. Because this skill is framed broadly around Word-like editing and productivity support, automatic activation raises the risk of overreach, accidental invocation on sensitive enterprise documents, and user confusion about what specialized behavior has been engaged.
This file is presented as a Chinese README, but core descriptive and trigger content is written in English, which can impose a language preference without explicit user opt-in. The file does not indicate that users may choose their preferred language or locale.
The example invocation phrases are written in broad everyday language like 'Help me' and 'I need a practical workflow', which are common across many unrelated requests. Even though they mention the general demand statement, they still encourage permissive matching behavior and can train downstream systems or authors to rely on low-specificity triggers. This primarily increases accidental invocation and user confusion rather than direct code-execution risk.
No suspicious patterns detected.
