Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording, but it does not install code, request credentials, persist state, or perform hidden actions.
Before installing, be aware that the skill may be selected for broad search or workflow wording more often than intended. It appears safe as a guidance-only helper, but users who maintain skill routing should narrow its triggers to explicit multi-search-engine workflow requests.
The manifest description is broad enough to apply to a wide range of user requests, including generic asks for 'practical workflow, artifact, checklist, analysis, or implementation support.' This ambiguity can cause the skill to be selected in contexts unrelated to multi-search workflows, increasing misrouting risk and making downstream behavior less predictable and less safe.
The trigger list includes extremely generic terms like 'multi', 'search', 'engine', 'global', and 'supports', which can match many unrelated conversations and cause accidental skill activation. Overbroad activation increases the chance that the skill intercepts requests outside its intended scope, leading to prompt-routing errors, user confusion, and unsafe delegation to an irrelevant workflow.
The trigger keywords include very broad everyday terms such as "multi", "search", "engine", and "global", which are likely to match many unrelated user requests. This can cause the skill to activate outside its intended scope, leading to workflow hijacking, irrelevant assistance, and reduced predictability of agent behavior.
The trigger sentence is overly broad and can match ordinary user language unrelated to this specific skill, causing accidental activation or routing. In an agent system, ambiguous activation expands the skill's execution surface and can lead to unintended workflow invocation, irrelevant outputs, or interference with more appropriate skills.
The trigger phrases are broad and keyword-heavy, which can cause the skill to activate for loosely related requests such as generic 'search', 'engine', or 'integration' tasks. In an agent ecosystem, overbroad routing increases the chance of unintended invocation, causing the wrong workflow to handle user input and potentially exposing data, producing unsafe actions, or bypassing more appropriate safeguards.
This file presents the user-facing documentation entirely in Chinese, while also describing usage and trigger phrases without stating that language selection is optional. That creates a locale constraint in the natural-language interface with no visible opt-in or justification in this file.
The trigger phrases include very generic terms such as 'multi', 'search', and 'engine', plus broad natural-language examples that can match many unrelated requests. This can cause unintended skill activation, leading the agent to route user tasks into this workflow when the user did not intend to invoke it, which is a prompt-routing and safety-boundary issue rather than just a usability bug.
The example trigger sentences are so generic and repetitive that they do little to constrain when the skill should fire, and they normalize activation on loosely related phrasing. Weak examples can reinforce poor routing heuristics and contribute to false positives, especially when combined with the already broad manifest and keyword list.
该技能文件为 SKILL.zh-CN.md，全文内容固定为简体中文语境，但文中没有说明这是可选语言版本、也没有给予用户语言偏好选择。按规则，若技能强制特定语言或区域设置而无用户 opt-in，属于自然语言层面的策略问题。
The description says to use the skill when users mention broad categories like work-productivity, multi, search, or engine, or whenever they need practical workflow or analysis support for the requirement. These boundaries are vague and expansive, making it easy for unrelated tasks to be routed into this skill and increasing the chance of incorrect or overbroad invocation.
The default prompt is overly broad and phrased in common language, which increases the chance the skill will be invoked in contexts the user did not explicitly intend. In combination with a productivity-oriented description and generic trigger terms, this can cause prompt-routing confusion, unexpected execution paths, or accidental exposure of the skill in unrelated conversations.
Enabling implicit invocation without meaningful constraints allows the platform to auto-select this skill based on loose semantic matches rather than explicit user intent. Because the skill targets broad 'work productivity' and 'search workflow' use cases, the context makes accidental invocation more likely and can lead to unintended actions, irrelevant guidance, or interference with other skills.
The activation examples are ambiguous and do not clearly distinguish this skill from generic productivity, workflow, or search-related requests. This makes misrouting more likely, especially in multi-skill agent environments where overlapping triggers can cause the wrong skill to activate and process user input under incorrect assumptions.
