Vague Triggers
Medium
- Confidence
- 88% confidence
- Finding
- The trigger list is unusually broad and includes generic terms like "helicopter," "friendly fire," "NVA," and even installation-related phrases, which can cause the skill to activate outside a clear request for this specific content. Over-broad activation can hijack unrelated conversations, override user intent, and increase prompt-injection surface by forcing the model into this skill when the context does not warrant it.
