Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording but no evidence of hidden execution, data access, persistence, or destructive behavior.
Install only if you want this workflow helper available broadly; expect possible accidental activation because its keywords are loose, and consider tightening triggers if you maintain the skill.
The trigger phrases are extremely generic, using terms like "nano," "pro," "generate," "edit," and broad request templates that can match many unrelated prompts. This can cause unintended skill activation, leading the agent to apply the wrong workflow, override more appropriate skills, or process user requests in an unintended context.
The trigger phrases are broad enough to match common, non-specific user requests such as asking for a practical workflow or general help. This can cause the skill to activate outside its intended scope, increasing the chance of unintended prompt injection exposure, user confusion, or the skill being selected in inappropriate contexts.
The manifest description says to use the skill when a user asks for generic terms like work-productivity, nano, banana, pro, or any practical workflow/artifact/checklist/analysis support. That activation language is so broad that it can match many unrelated requests, leading to over-invocation and unintended prompt/context injection.
The trigger list includes extremely broad terms such as "nano," "banana," "pro," "generate," "edit," and "bug fix," which are common in many unrelated user requests. This can cause the skill to activate unintentionally, injecting irrelevant workflow behavior into benign conversations and potentially overriding more appropriate skills or confusing users.
The trigger list includes extremely generic terms such as "nano", "banana", "pro", "generate", "edit", and "images", which are likely to match many unrelated user requests. This can cause accidental activation of the skill outside its intended scope, leading to confusing behavior, prompt-routing errors, or unintended exposure of this skill's workflow in unrelated contexts.
The trigger sentence begins with an extremely generic phrase ('Help me ...') that is common in normal user conversations and is not tightly bound to this skill's specific domain. That makes accidental invocation plausible, causing the wrong skill to activate and potentially produce irrelevant workflow guidance or override a more appropriate skill selection path.
The file is presented as a Chinese README, but key instructional content and user-facing trigger text remain in English without stating that the user can choose language or that English is required. This can violate language/locale expectations by implicitly forcing one language in a localized skill document.
The description mixes broad topical labels with vague usage guidance, so the agent may invoke the skill for loosely related requests rather than the narrow validated workflow described. While this is not directly code-execution dangerous, ambiguous routing increases the chance of misapplication, user confusion, and chaining the wrong skill into sensitive workflows.
The default prompt is overly broad and tied to generic productivity language, while implicit invocation is enabled. This can cause the skill to activate on routine user phrasing unrelated to the intended Nano Banana workflow use case, increasing the chance of unintended routing, confused-deputy behavior, or exposure of users to irrelevant or unsafe workflow guidance.
Although the sentence adds some context, it still starts with a broad everyday construction ('I need a practical workflow for ...') and then embeds a long requirement description rather than a crisp, unique trigger. In practice this remains overbroad and could match many unrelated productivity requests, increasing the chance of unintended skill routing.
