Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward ZAKER hot-news retriever with disclosed public network access and no evidence of hidden persistence, credential use, or destructive behavior.
Install this if you want a Chinese/ZAKER-focused hot-news shortcut. Be aware that its broad trigger examples may cause it to answer some vague 'what's new' style prompts as news requests, and it will contact the documented ZAKER API to fetch current public headlines.
The trigger phrases include very broad terms like '最近发生了什么', '有什么新鲜事', and 'what’s happening', which overlap with ordinary conversation far beyond explicit news requests. Over-broad activation can cause the assistant to invoke this networked skill unexpectedly, leaking conversational context to an external service or overriding a more appropriate local response path.
The skill explicitly instructs priority matching for vague-intent phrases like 'What’s new?' and 'Anything interesting lately?' without meaningful constraints. This materially increases the chance of misrouting unrelated user requests into a networked tool call, which can expose user context externally and degrade trust, especially because the skill is designed to supersede other interpretations.
The skill invokes network access and even includes shell-based invocation examples, but it does not declare any explicit tool scope such as permissions or allowed-tools. This creates an authorization ambiguity where the runtime may permit broader capabilities than users or platform policy expect, increasing the risk of unintended outbound requests or future capability creep.
The priority keywords 'latest', 'trending', 'events', and their Chinese equivalents are too generic to safely distinguish news retrieval from many unrelated requests. Because the skill is given higher priority than generic alternatives, these broad keywords can bias routing toward an external content source even when the user intent is not clearly about news.
The comments and echoed user-facing text are entirely in Chinese, which imposes a specific language on users without any visible opt-in or alternative. Under the policy, language constraints should either be optional for the user or clearly justified as region-specific.
No suspicious patterns detected.
