Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local AI learning journal that writes Markdown notes and summaries, with no executable code or network behavior, but users should know it may activate on casual AI-related comments.
Install only if you want a local, file-backed AI learning journal. Be aware that AI-related remarks may be organized into records under the skill's records directory; ask the agent not to save when you only want a transient conversation.
The skill explicitly states it should trigger on very broad AI-related topics, including casual remarks like 'tried Claude today' or 'learned a new prompting trick.' That can cause unintended invocation and unexpected handling of user content, especially because the skill also persists records to disk, increasing privacy and consent risk.
The skill instructs the agent to create and modify persistent Markdown files under a local records directory, but it does not require explicit disclosure or user consent before storing data. Users may casually mention experiences without realizing those details will be written to disk, creating privacy and surprise-retention risks.
The trigger examples for 'learning guidance' include generic phrases like '这个怎么用比较好？' and '有什么最佳实践吗？', which are common across many unrelated contexts. This ambiguity can cause the skill to activate outside its intended scope and steer conversations into file-backed journaling or AI-learning guidance when the user did not ask for that.
All user-facing instructions and invocation examples are presented only in Chinese, and the document does not indicate that users may choose another language. This can violate a language-choice policy when a skill implicitly requires a specific language without opt-in or justification.
The listed trigger phrases are broad, natural-language requests such as reviewing learning records or asking for a study plan. In a shared assistant environment, these phrases can unintentionally activate the skill during ordinary conversation, causing unexpected behavior such as modifying or surfacing journal content without clear user intent.
The README content is entirely in Chinese and does not indicate that the skill is region-specific or that users can opt into another language. Under the policy for natural-language violations, forcing a specific language without user choice can be a locale-policy issue.
No suspicious patterns detected.
