Security audit
Security checks for vulnerabilities and agentic risk
This is a simple workflow-helper skill with overly broad activation wording, but it does not contain code, hidden behavior, credential access, persistence, or destructive actions.
Installers should be aware that this skill may activate for generic words like nano, pro, generate, or edit. Consider narrowing its triggers or invoking it explicitly when you want this specific workflow helper.
The trigger phrases are broad, generic, and include common terms such as "nano," "pro," "generate," and "bug fix," which can cause the skill to activate for unrelated user requests. In an agent ecosystem, unintended invocation can route user tasks into the wrong workflow, causing confused delegation, irrelevant outputs, and possible bypass of more appropriate safeguards or specialized skills.
The manifest description says to use the skill when a user asks for broad terms like 'work-productivity', 'nano', 'banana', or 'pro', or whenever they need a 'practical workflow, artifact, checklist, analysis, or implementation support.' This is so expansive that it can match a wide range of ordinary requests, making accidental invocation likely and weakening skill-selection integrity.
The trigger keywords are extremely broad, including generic terms like 'nano', 'banana', 'pro', 'generate', and 'edit', which can cause this skill to activate for many unrelated user requests. Over-broad activation increases the chance of unintended routing, where users receive this skill's guidance instead of a more appropriate or safer skill, reducing predictability and creating opportunities for prompt-surface interference.
The trigger list includes very common standalone terms like "nano", "banana", "pro", "generate", and "edit", which are likely to match many unrelated prompts. This can cause the skill to activate outside its intended scope, leading to incorrect routing, user confusion, and increased exposure to any unsafe or low-quality behavior embedded in the skill.
The trigger sentence is excessively broad and resembles normal user language, which can cause the skill to activate in situations where the user did not explicitly intend to invoke it. In an agent ecosystem, unintended activation can steer workflows, inject irrelevant instructions, or override more appropriate skills, reducing reliability and potentially creating a path for prompt-routing abuse.
The trigger phrases are broad enough to match ordinary words like 'work-productivity', 'generate', 'edit', 'images', and 'bug fix', which can cause the skill to activate in many unrelated conversations. That increases the chance of unintended routing, confusing behavior, and accidental invocation of workflow guidance when the user did not request this specific skill.
The example trigger phrases model invocation patterns based on ordinary language such as 'Help me' and 'I need a practical workflow' without meaningful constraints. These examples can reinforce overly permissive matching behavior and make it more likely that unrelated requests are interpreted as calls to this skill.
The description says to use the skill when users mention broad terms or need generic artifacts like workflows, checklists, analysis, or implementation support for the requirement, without clear boundaries. This ambiguity makes over-activation likely, so the skill may intercept requests better handled by other skills and operate in contexts the author did not adequately constrain.
The default prompt is broad, truncated, and repeats generic demand-language rather than a narrowly scoped task trigger, which increases the chance of unintended or overly broad invocation. Combined with allow_implicit_invocation: true, the skill may activate in contexts where the user did not clearly request it, causing inappropriate workflow assistance, prompt injection surface expansion, or confusing agent behavior.
The invocation guidance uses loose keywords and broad trigger sentences without clear boundaries, increasing the chance of accidental or inappropriate skill selection. Because this skill is positioned as a general workflow helper, ambiguous routing can make the problem more dangerous by causing it to capture a wide range of unrelated requests and influence downstream agent behavior.
The file is presented as a Chinese README, but the key usage content and trigger phrases are written in English only. This can amount to forcing a specific language for activation or use without explicit user opt-in or a documented language choice.
