Security audit
Security checks for vulnerabilities and agentic risk
This is a simple brainstorming prompt skill with no code execution, credential access, data access, persistence, or hidden behavior.
Installers should expect this skill to make the agent more proactive during brainstorming. If you prefer the agent to ask before ideating or to avoid inferred modes, review or tighten the activation wording, but no security-sensitive behavior was found.
The skill says 'When user asks for ideas, generate immediately' and relies on missing context being clarified later, which is a broad activation condition for a markdown skill. While examples are provided later, the file does not clearly define exclusion conditions or narrow trigger scope, so common phrases could unintentionally invoke this skill.
The phrase "Detect from context" creates a broad, implicit activation condition that can cause the skill to engage without a clear user request. In a brainstorming skill this is not directly dangerous, but it can lead to unintended invocation, mode confusion, and reduced user control over how the agent responds.
The instruction "Use when stuck or need more variety" is broad and subjective, making it unclear when this skill should activate versus when it should not. Without clearer scope or exclusion conditions, the prompt set could be invoked in many ordinary situations.
The phrase "When user seems stuck, offer" relies on an ambiguous assessment of the user's state and does not specify concrete signals or boundaries. This can cause unintended invocation because many normal pauses or short requests could be interpreted as being stuck.
No suspicious patterns detected.
