Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording but no hidden code, credential use, persistence, or destructive behavior.
Install only if you want a broad Nano Banana Pro workflow helper, and be aware it may activate on generic image, generation, editing, or bug-fix requests unless the host router applies stricter matching. The package itself does not appear to run code or access sensitive data.
The skill description is so broad that it can match many ordinary user requests unrelated to a narrowly scoped Nano Banana workflow. Overbroad activation increases the chance this skill is invoked unintentionally, causing prompt hijacking of benign tasks, misrouting, or inappropriate use of workflow instructions in unrelated contexts.
The trigger keywords include highly generic terms such as 'nano', 'banana', 'pro', 'generate', 'edit', 'images', and 'bug fix', which are likely to appear in many unrelated conversations. This makes accidental activation very likely and can let the skill intercept broad classes of requests it was never intended to handle.
The description says to use the skill whenever a user mentions broad terms or needs practical workflow, checklist, analysis, or implementation support for the requirement, which is too open-ended and lacks clear boundaries. This can cause accidental invocation across many unrelated tasks, leading to prompt-scope confusion and unintended skill chaining.
The trigger list includes extremely broad everyday terms such as "nano", "banana", and especially "pro", which can cause this skill to activate for unrelated user requests. Over-broad activation increases the chance that the agent applies the wrong workflow, injects irrelevant instructions into another task, or surfaces unintended behavior without the user explicitly opting in.
The trigger sentence is extremely broad and reads like ordinary user language, which can cause the skill to activate in situations far outside its intended scope. In an agent environment, unintended invocation can route unrelated requests into this workflow, producing incorrect actions, confusing behavior, or unsafe handling of tasks that should have been processed by a more appropriate skill.
The trigger phrases are very broad and include generic terms such as 'nano', 'banana', 'pro', 'generate', and 'bug fix', which can cause the skill to activate in unrelated contexts. This creates prompt-routing ambiguity and increases the chance that users are steered into this workflow when they intended something else, potentially causing incorrect actions, disclosure of unrelated context, or unsafe task execution under the wrong skill.
The listed trigger keywords and example phrases are extremely generic, including common terms like 'nano', 'pro', 'generate', 'edit', and 'images'. This can cause the skill to activate during ordinary conversations unrelated to this specific workflow, leading to unintended invocation, user confusion, and possible routing of sensitive or irrelevant tasks into the skill's instructions.
The example trigger phrases are vague, truncated, and do not define a reliable boundary for when the skill should or should not run. Ambiguous examples reinforce overmatching behavior and make it harder for maintainers or routers to distinguish intended activation from ordinary user language.
该技能文件为 SKILL.zh-CN.md，全文内容以简体中文呈现，但文件中未说明这是用户可选择的本地化版本，也未提供语言切换或按用户偏好响应的说明。按规则，若技能强制特定语言而无用户选择或合理限定，属于语言/区域策略风险。
The default prompt contains a long, generic trigger phrase built from common productivity terms and broad user intents, which increases the chance of implicit invocation during ordinary conversation. Because the policy also enables implicit invocation, unrelated user requests about workflows, checklists, analysis, or implementation support could accidentally activate this skill and inject its behavior into sessions without clear user intent.
The activation guidance is ambiguous because it mixes broad keywords ('nano', 'pro', 'generate', 'edit', 'images', 'bug fix') with generic trigger templates that do not require strong contextual matching. This increases the chance of accidental or priority-skewed skill selection, especially in multi-skill agent systems where overlapping language can cause the wrong workflow to run.
