Security audit
Security checks for vulnerabilities and agentic risk
This is a low-risk workflow-advice skill, but its activation wording is overly broad and may appear in unrelated chats.
Install only if you are comfortable with the skill being auto-selected for loosely related words; users or maintainers should narrow the trigger list to the full skill name or specific Nano Banana Pro workflow phrases.
The trigger keywords and example phrasing are broad enough to match common words like 'work-productivity', 'nano', 'banana', 'pro', 'generate', and 'edit', which can cause the skill to activate in unrelated conversations. Over-broad activation can hijack user intent, invoke the wrong workflow, and increase the chance that downstream instructions are applied outside their intended context.
The trigger phrases are broad and contain common terms such as 'help me', 'I need', 'workflow', and generic product words, which can cause the skill to activate in contexts the user did not intend. In an agent ecosystem, unintended invocation can route user requests into the wrong skill, causing prompt hijacking opportunities, workflow confusion, or inappropriate handling of user data and actions.
The description says to use the skill whenever a user asks for broad categories like work-productivity, practical workflows, artifacts, checklists, analyses, or implementation support, which are common across many unrelated tasks. This ambiguous invocation guidance can cause over-selection of the skill, reducing routing accuracy and increasing the risk of irrelevant or inappropriate outputs being presented as authoritative.
The trigger list includes extremely broad everyday terms such as "nano," "banana," "pro," "generate," and "edit," which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that the wrong skill is invoked, leading to misleading guidance, workflow confusion, or unintended handling of user requests outside the skill's intended scope.
The trigger keywords are excessively broad, including generic terms like "nano", "banana", "pro", "generate", and "edit", which can cause the skill to activate for unrelated user requests. This creates a prompt-routing and least-surprise problem: users may be routed into this skill unintentionally, causing irrelevant guidance, conflicting instructions, or unsafe workflow substitution in contexts the skill was not designed for.
The trigger sentence is so broad and malformed that it can cause the skill to activate for generic user requests unrelated to the intended Nano Banana workflow use case. Overbroad activation increases the chance of incorrect routing, prompt hijacking surface, and unintended handling of tasks outside the skill’s reviewed scope.
The activation examples are ambiguous, truncated, and do not define when the skill should or should not trigger, making accidental invocation more likely. In an agent environment, unclear trigger scope can cause the skill to supersede more appropriate skills or inject workflow behavior into unrelated tasks, reducing reliability and potentially affecting safety-sensitive requests.
该技能文件整体为 zh-CN 中文版本，且内容中未见向用户提供语言/地区选择或说明这是用户自选的本地化变体。根据规则，若技能在自然语言层面强制特定语言而没有用户选择或合理限定，可能构成语言/地区策略问题。
The description states the skill should be used when users mention broad categories or need "practical workflow, artifact, checklist, analysis, or implementation support," without clearly defining boundaries. This ambiguity increases the chance of accidental invocation and overlapping behavior with many unrelated skills, reducing predictability and potentially causing the wrong instructions or outputs to be applied.
The default prompt activates on very broad terms and lacks clear boundaries on when the skill should be invoked. This can cause unintended routing of unrelated user requests into the skill, increasing the chance of prompt-context confusion, overcollection of user data, or unsafe execution of the skill in situations the user did not intend.
Enabling implicit invocation without strict activation boundaries allows the platform to auto-select this skill based on loosely related user input. In combination with the broad skill description, this raises the risk of accidental invocation, unintended influence over conversations, and misuse in contexts where the skill is not appropriate.
The invocation guidance uses broad keywords and request patterns such as 'nano', 'pro', 'generate', and 'edit', which are common across many unrelated tasks. This makes the skill susceptible to accidental or overly aggressive invocation, potentially causing it to intercept prompts beyond its intended scope and increasing the risk of unsafe or irrelevant behavior in downstream workflows.
A file named as Simplified Chinese documentation contains substantial English prose in the core requirements section, which can impose an unintended language experience on users expecting Chinese content. The file does not indicate that bilingual content is intentional or offer a language/locale choice.
No suspicious patterns detected.
