Vague Triggers
Medium
- Confidence
- 95% confidence
- Finding
- The skill defines activation on a very wide set of common phrases and entities, including generic topics like AI safety, Twitter/X, electric vehicles, and even onboarding-style statements such as a user saying they just installed the skill or do not know how to start. This can cause unsolicited invocation and instruction injection into unrelated conversations, increasing the chance of confusing users, overriding normal assistant behavior, or surfacing irrelevant book-specific guidance when it was not requested.
