Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only skill for improving error messages, with broad trigger wording but no hidden execution, persistence, credential access, or destructive behavior.
Installers should be aware that this skill may be invoked for general debugging or support prompts because its trigger wording is broad. It appears safe for its intended use, but users may prefer narrower triggers if they only want it used for explicit error-message rewriting or troubleshooting-copy tasks.
The manifest description includes very broad use conditions such as work-productivity, debugging, user feedback, and support, which are common across many unrelated conversations. This can cause accidental or overbroad invocation of the skill, increasing the chance that it influences prompts outside its intended scope and creates confusing or unsafe task routing.
The trigger phrases are broad natural-language requests such as 'Help me...' and 'I need a practical workflow...', which can match many ordinary user prompts and cause the skill to activate unintentionally. In an agent environment, unintended invocation can override more appropriate handling, inject irrelevant workflow constraints, or expand the skill’s influence beyond the user’s explicit intent.
The file is a Chinese-localized README and presents all user-facing guidance in Chinese without any explicit opt-in or language choice within the document. Per the policy, forcing a specific language can be a natural-language policy violation unless the locale constraint is clearly documented and justified.
The trigger phrases are extremely broad and map to common support and productivity requests, which can cause the skill to activate outside its intended scope. Overbroad activation increases the chance of prompt/skill hijacking, user confusion, and inappropriate routing of benign requests into this workflow without clear user intent.
The trigger keywords are highly generic and broad, especially debugging, support, and troubleshooting, which can match a large portion of normal user requests. Overly general triggers raise the risk of unintended activation, causing the skill to intercept tasks it was not designed to handle and potentially degrade system behavior or produce irrelevant guidance.
The example trigger phrases use everyday language like help me and I need a practical workflow, which are common conversational patterns and therefore prone to accidental matching. Although the skill itself is not high-risk, these examples reinforce imprecise routing behavior and make over-invocation more likely.
The skill description defines applicability in broad, ambiguous terms and does not clearly state exclusion criteria. Ambiguous routing logic increases the chance of mis-selection, which can mislead users, interfere with other skills, and reduce trust in system behavior.
The trigger keywords are very broad and overlap with common conversational requests such as debugging, support, and user feedback. This can cause unintended activation of the skill in unrelated contexts, leading the agent to apply the wrong workflow, produce irrelevant outputs, or override more appropriate skills.
The default prompt is phrased so broadly that it can match many ordinary requests about work productivity, debugging, support, or error messages, which increases the chance of implicit or unintended invocation. Because implicit invocation is enabled, the skill may activate in contexts the user did not explicitly intend, causing prompt-scope expansion, unexpected behavior, or routing of sensitive troubleshooting content into this skill.
The trigger scope is ambiguous because it does not define boundaries for when this skill should not run, and it overlaps with many common support and productivity requests. In agent environments, ambiguous routing can lead to unauthorized or unintended skill invocation, causing the model to apply the wrong workflow and potentially mishandle sensitive debugging or support content.
The trigger sentence is so broad and natural-language-like that ordinary user requests could unintentionally activate this skill. That can cause incorrect routing, prompt injection surface expansion, or the skill being invoked outside its intended scope for generic help requests about work or support.
The README declares separate English and Chinese instruction files, but it does not state how language is selected or whether the user can choose their preferred locale. This can create a locale-policy concern if the skill defaults to a language without explicit opt-in.
该技能文件整体以中文呈现，文件名为 SKILL.zh-CN.md，但正文未说明这是可选语言版本，也未向用户提供语言/地区选择。按规则，若技能在自然语言层面强制特定语言而没有用户选择或明确说明，可能构成语言/地区策略问题。
