The system processes product reviews using tool use with a defined schema: rating (integer 1-5), pros (string array), cons (string array), and overall_sentiment (enum: positive, negative, neutral). Testing reveals two issues with brief or ambiguous reviews (~20% of the dataset): (1) for reviews like "Great product!", Claude fabricates specific pros and cons rather than indicating information isn't explicitly stated, and (2) for sarcastic reviews like "Well that was... interesting", Claude picks sentiment arbitrarily since there's no option for ambiguous cases. Which modification best addresses both issues?