Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-style workflow helper with overly broad activation wording, but it does not install code, request credentials, persist itself, or perform hidden actions.
Install this only if you want a general Agent Browser workflow helper. Be aware that its activation wording is loose, so it may be invoked for ordinary browser, automation, CLI, or bug-fix requests unless your agent/router applies tighter selection rules.
The trigger list is broad enough to match many ordinary requests such as 'browser', 'automation', or 'cli', which can cause this skill to activate outside its intended scope. Over-broad activation increases the chance of prompt injection surface expansion, incorrect skill routing, and unintended influence over unrelated tasks.
The trigger phrases are broad, generic, and partially malformed, making it easier for the skill to be invoked in contexts the user did not clearly intend. In an automation/browser workflow skill, unintended invocation can cause the agent to apply the wrong workflow, produce irrelevant actions, or initiate browser/headless guidance in a context where it is not appropriate.
The trigger phrases are broad enough to activate on common terms like browser, automation, cli, or bug fix, which can cause the skill to run in contexts the user did not intend. In an agent workflow, overbroad activation increases the chance of unintended execution, incorrect tool use, or the skill influencing unrelated tasks.
The example trigger phrases use very common helper language such as 'Help me' and 'I need a practical workflow', making it easier for normal conversations to spuriously match this skill. This can cause unintended invocation and broaden exposure of the skill's instructions in contexts where they are irrelevant or risky.
技能描述将适用范围定义为 work-productivity, agent-browser, browser, headless, automation 以及几乎任意“实用流程、产物、检查清单、分析或实现支持”，边界过宽，缺少排除条件。这样的范围膨胀会增加路由歧义，使该技能在与其核心任务无关的请求上被调用，放大错误建议或不恰当自动化输出的风险。
触发关键词包含 browser、automation、cli、optimized、bug fix 等高频通用词，极易与大量普通任务重叠，导致技能在非目标场景被误触发。误触发后会把用户请求导向该技能的预设流程，可能造成错误自动化建议、越权式上下文扩展或掩盖更合适的专用技能。
The default prompt uses very broad, everyday trigger terms such as work-productivity, browser, headless, automation, workflow, checklist, analysis, and implementation support. This creates an overbroad matching surface that can cause the skill to be invoked for many unrelated requests, amplifying prompt-routing risk and making unintended tool or workflow guidance more likely.
allow_implicit_invocation: true permits the skill to be activated without an explicit user request, which increases the chance that it runs in contexts the user did not intend. In this skill, the broad productivity/browser framing makes accidental or overly permissive activation more likely, potentially causing the agent to apply browser-automation guidance or workflow behavior outside narrowly appropriate scenarios.
The invocation guidance is underspecified and does not define strong boundaries for when the skill should be selected, increasing the chance of accidental or opportunistic activation. Because this skill is positioned around browser-style workflows and automation, ambiguous routing is more dangerous than in a purely informational skill: it may lead to unintended actions, data exposure to tools, or execution of the wrong workflow.
The trigger sentence is broad and resembles normal user phrasing, which can cause the skill to activate in conversations that are only loosely related to the intended workflow. In an agent environment, unintended invocation can route user input into the wrong automation path, causing confusing behavior, unnecessary tool use, or unsafe execution in a browser/headless context.
The README explicitly labels separate English and Chinese instruction and guide files, but does not state whether the skill adapts to the user's preferred language or requires opt-in. Because language handling is part of organizational policy review, this can be read as a locale behavior that is undocumented from the user's perspective.
The file presents the skill instructions and usage entirely in Chinese while the skill itself is not described as China-specific or locale-restricted. Under the language/locale policy, forcing a specific language without user opt-in can be a natural-language policy violation.
示例触发句使用 Help me、I need a practical workflow for 这类非常常见的自然语言前缀，并拼接宽泛需求文本，容易让普通求助语句被错误解释为该技能调用信号。虽然单条影响较低，但会进一步强化前述宽范围匹配问题，增加误触发概率。
No suspicious patterns detected.
