Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed A-share capital-flow analysis skill with financial-advice risk, but it does not contain hidden code, credential use, trading execution, or destructive behavior.
Install only if you want Chinese-language A-share capital-flow analysis. Do not treat its buy, sell, price, or position-sizing suggestions as professional financial advice, and review any scheduled reports manually before acting on them.
The example trigger phrases are generic financial-analysis requests that could easily overlap with ordinary user conversation, increasing the chance the skill is invoked unintentionally. Because the README provides no explicit activation boundary, namespace, or scope limitation, the skill may activate on broad stock-market queries and influence responses without clear user intent.
The skill explicitly offers trading decisions and concrete buy/sell strategy support but does not include any warning that outputs may be inaccurate, incomplete, or unsuitable as financial advice. In a financial context, users may over-rely on the model’s recommendations and make harmful investment decisions without understanding the risks or limitations.
The scheduled cron reports automatically generate operation suggestions at fixed times, which increases the chance that unattended or habitual consumption of the output will influence real financial decisions without active user review. Because no warning or approval gate is described, the automation amplifies the risk of users acting on unvetted recommendations.
The trigger list is broad and consists of common stock/trading terms such as '股票', '行情', and '交易', which are likely to match many normal user conversations beyond the skill's intended scope. This can cause the skill to activate unexpectedly, increasing the chance of inappropriate interception of financial discussions and unintended influence over investment-related interactions.
The natural-language instructions and example invocations are exclusively in Chinese, which can amount to forcing a specific language without user opt-in. There is no indication that the skill supports other languages or that the Chinese-only scope is an explicit, justified locale constraint.
The file presents all user-facing instructions and invocation phrases in Chinese only, with no indication that users can select another language or that the locale restriction is intentional. The policy requires flagging forced language or locale behavior when there is no user opt-in or documented justification.
No suspicious patterns detected.
