Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward Chinese text proofreading and polishing aid with no hidden execution, network access, credential use, or privileged behavior.
Install this if you want an agent to polish or proofread supplied Chinese text. Be aware that broad phrases like "check this text" may invoke it during ordinary editing requests, so review important edits before sending or publishing.
The one-line description uses broad, natural phrasing such as '发出去之前过一遍' and promises to fix multiple common writing issues in a single pass, without clear activation boundaries. In agent ecosystems, this can cause the skill to be invoked for a wide range of generic writing or editing requests, increasing unintended routing and prompt-scope overlap with other skills.
The README presents all user-facing instructions and invocation examples exclusively in Chinese, which can amount to a language policy constraint if the skill expects interaction in that language by default. There is no opt-in language choice or statement that the skill is intentionally limited to a Chinese-language context.
The invocation examples are generic phrases like '检查这段文字' and '帮我校对一下', which overlap heavily with ordinary user requests rather than clearly scoped skill activation. This can cause the skill to trigger unintentionally in broad contexts, leading to over-application of text rewriting behavior and reducing user control over whether a skill is invoked.
The description states the skill improves "中文文本" and elsewhere prefers normalization to all-Chinese or all-English, but it does not clearly present this as an opt-in language constraint. This can violate language/locale policy when users are not given a choice or informed that the skill is intentionally Chinese-specific.
The invocation phrases are extremely broad and map to common conversational requests such as “改一改” and “检查文字,” which can cause the skill to trigger in situations the user did not clearly intend. Over-broad activation can lead to unintended handling or transformation of user content, increasing the chance of misrouting prompts and interfering with other more appropriate skills.
The target-audience section is extremely broad ('任何不想发出错别字的人'), which effectively maps to almost any user composing text. This broad positioning can make the skill appear applicable to generic writing interactions, increasing accidental activation and reducing precision in multi-skill selection.
No suspicious patterns detected.
