Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad auto-invocation wording, but it does not install code, persist, access credentials, or perform hidden actions.
Installers should be aware that this skill may activate too broadly because of generic trigger words and implicit invocation. Prefer explicit invocation by skill name, and the publisher should narrow the trigger language before relying on automatic routing.
The skill description says to use the skill when a user asks for very broad terms like 'work-productivity', 'security', or 'needs a practical workflow', which are common across many unrelated requests. This can cause unintended activation and prompt/context injection into tasks outside the intended scope, reducing routing precision and potentially influencing outputs in inappropriate contexts.
The keyword list includes ambiguous common words like 'security', 'gate', 'every', 'must', 'pass', and 'before', which are likely to appear in many ordinary prompts. In an agent ecosystem, such broad triggers can hijack routing, over-invoke this skill, and introduce irrelevant or conflicting instructions into unrelated tasks.
The trigger keyword list includes very broad everyday terms such as "security", "every", "must", "pass", "before", and "activate", which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance of unintended instruction injection into workflows, misrouting user requests, and unsafe automation being applied outside its intended context.
The trigger guidance includes very broad everyday terms and malformed example phrasing, which can cause the skill to activate in contexts unrelated to the intended security workflow. Unintended activation is dangerous because it can hijack routing, override more appropriate skills, and cause users' unrelated requests to be processed under a security-oriented workflow without clear consent.
The keyword list and activation guidance are ambiguous and include common words such as 'security', 'gate', 'every', 'must', and 'before', which are likely to appear in everyday conversation. In an agent ecosystem, this can cause accidental skill selection or routing conflicts, making the skill activate outside its intended scope and reducing reliability and safety of downstream behavior.
The trigger sentences are broad, generic prompts that can easily match ordinary user requests unrelated to this specific skill. That increases the chance of unintended activation, causing the agent to invoke security-workflow behavior in the wrong context and potentially interfere with normal task routing or apply security guidance where it was not requested.
This README is presented entirely in Chinese and labels itself as the user-facing documentation for the skill, but it does not state that language selection is optional or limited to a China-specific use case. Under the policy, user-facing language constraints should be opt-in or clearly justified.
The trigger keywords and example phrases are extremely broad, including generic terms like "security," "every," "must," "pass," and "before," which can cause the skill to activate on many unrelated user requests. In an agent ecosystem, this increases the chance of unintended routing or shadowing of more appropriate skills, which can lead to incorrect task handling and unsafe automation decisions.
The example trigger sentences are malformed and mostly repeat marketing text rather than clearly showing activation conditions. Poorly delimited examples make it easier for trigger matchers or developers to misinterpret what should activate the skill, increasing accidental invocation and unreliable behavior.
The skill description defines invocation conditions too broadly, covering generic topics like work-productivity, security, checklists, analysis, or implementation support without clear boundaries. This makes the skill eligible for many unrelated requests, increasing accidental invocation and creating opportunities for context confusion or misuse in conversations that were not meant to use this skill.
The skill enables implicit invocation, but the trigger description is broad and vague, covering generic terms like work-productivity, security, workflow, checklist, and analysis. This can cause the agent platform to auto-select the skill in contexts beyond its intended scope, increasing the chance of unintended execution, prompt-scope confusion, or misuse in sensitive security-related conversations.
The invocation section mixes broad matching terms with malformed trigger examples, increasing the chance of accidental or overbroad activation. In an agent ecosystem, this can distort intent classification and route ordinary productivity or security-related prompts into this skill unexpectedly, reducing reliability and potentially interfering with safer or more appropriate handlers.
