Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed financial catalyst-calendar helper that searches for upcoming events, saves a report, and creates reminder notes, with no evidence of hidden code or harmful behavior.
Install this if you want a Chinese-language financial catalyst calendar that may run web/financial searches, write dated Markdown reports, and create reminder notes. Review or disable the reminder/report behavior if you only want ephemeral chat answers or if your market watchlist is sensitive.
The trigger phrases are broad, generic, and likely to match ordinary user requests about upcoming events, which can cause the skill to activate unintentionally. In this context, unintended activation is more concerning because the skill performs web searches and writes reports to disk, potentially causing unexpected external tool use and persistence without clear user intent.
The README states that the skill saves a markdown report to disk but does not clearly warn users that local persistence will occur. This can lead to silent storage of potentially sensitive queries, market interests, or generated research artifacts, especially if the user expected an ephemeral response only.
The activation rule “需要时间维度的事件前瞻时触发” is ambiguous because it does not define subject matter boundaries, so the skill may activate for unrelated planning, news, or general forecasting requests. In an agent environment, vague routing logic increases the chance of misfires and unexpected tool use.
The activation rule “需要时间维度的事件前瞻时触发” is ambiguous because it does not define subject matter boundaries, so the skill may activate for unrelated planning, news, or general forecasting requests. In an agent environment, vague routing logic increases the chance of misfires and unexpected tool use.
The skill instructs the agent to write a report file and create IMA reminder notes, including proactive reminders, without any explicit user consent or confirmation step. This creates an unauthorized side-effect risk: the agent may persist data or generate notifications the user did not request, which is especially problematic if outputs contain sensitive interests, watchlists, or trading-related context.
The README title, feature descriptions, trigger phrases, and output description are all written in Chinese, and there is no indication that users may choose another language or that the skill is intentionally limited to a Chinese-speaking context. This can conflict with language/locale policy expectations when a skill appears to force one language without opt-in.
All natural-language instructions and examples are presented in Chinese, and there is no statement that the skill is region-specific or that users may choose another language. This can violate language/locale policy when a skill implicitly forces one language without opt-in.
No suspicious patterns detected.
