Security audit
Security checks for vulnerabilities and agentic risk
This documentation-only local LLM advice skill has overly broad activation text, but no hidden execution, data access, persistence, or credential handling.
This skill appears safe to install from a security perspective. Review its broad trigger wording if you rely on automatic skill routing, because it may be invoked for loosely related local AI questions, but it does not appear to perform actions or access data on its own.
The trigger phrases are broad, repetitive, and partially truncated, which can cause the skill to activate outside the author's intended scope. In an agent environment, ambiguous routing can lead to the wrong skill being selected for user requests, producing irrelevant advice and increasing the chance that downstream instructions are applied in an unintended context.
The trigger phrases are broad enough to match ordinary user requests about local LLMs, workflows, or practical help, which can cause the skill to activate outside its narrowly intended scope. Over-broad routing increases the chance that users are funneled into this skill unexpectedly, potentially exposing them to irrelevant guidance, prompt collisions, or unsafe downstream behaviors if the skill later gains more powerful capabilities.
The skill description and activation guidance are broad enough to match many general AI/help requests, which can cause the skill to be invoked outside its intended niche. Over-broad routing is dangerous because it can misapply specialized instructions, increase unintended tool use, or suppress better-matched skills in unrelated contexts.
The example trigger uses a generic phrase pattern that could match ordinary user requests with only weak linkage to the skill’s actual scope. If a router learns from or relies on these examples, it may over-select this skill for ambiguous prompts, leading to scope creep and reduced safety or correctness in downstream handling.
文件名和正文整体均以中文呈现，但未说明这是可选语言版本、也未向用户提供语言选择或切换方式。对于面向广泛用户群体的技能，这可能构成未经用户明确选择就施加特定语言/locale 的策略问题。
The trigger keywords are broad and lack clear exclusion criteria, which can cause the skill to activate outside its intended scope. In an agent system, overbroad invocation can route unrelated user requests into this skill, causing inappropriate guidance, reduced reliability, and possible policy bypass through unintended tool selection.
The example trigger phrases use generic language such as 'Help me' and 'I need a practical workflow' without boundaries, which reinforces ambiguous routing behavior. If the orchestrator relies on examples for activation, the skill may be selected for loosely related requests, increasing the chance of incorrect or unsafe automation paths.
The default prompt contains a very broad invocation phrase tied to common concepts like local LLMs, software, and implementation help, which can cause the skill to be auto-invoked in ordinary user conversations that were not meant to activate it. This increases the risk of unintended routing, prompt-context injection into unrelated tasks, and overcollection or misuse of user intent within broader conversations.
The trigger section lacks precise scope boundaries, so the skill may be invoked for loosely related requests without confirming that the user actually wants local-LLM setup advice. This is dangerous because ambiguous activation expands the skill's effective authority, increasing accidental invocation, prompt-routing errors, and unintended handling of requests beyond the documented requirement.
The trigger sentence is so broad and malformed that it can overlap with ordinary user phrasing, causing the skill to activate outside its intended scope. In an agent environment, over-broad activation can route unrelated requests into this skill, producing unintended guidance and increasing the chance of policy bypass or unsafe handling through misclassification.
This Chinese-language README contains core descriptive content in English, which can impose a language expectation without explicit user opt-in or a documented language-selection policy. The file does reference separate English and Chinese documents later, but this section itself does not clearly present a language choice to the reader.
No suspicious patterns detected.
