Vague Triggers
Medium
- Confidence
- 94% confidence
- Finding
- The invocation examples are very generic phrases like 'what's in this image?' and 'describe what you see,' which can overlap with normal conversation and cause the skill to trigger in broad, unintended contexts. In a multimodal assistant, this increases the chance of accidental routing of user content, including screenshots with sensitive information, into this skill without clear user awareness.
