conductor-checkpoint-msg_01ELC9AyZZGry9tN4XKrwEM6

This commit is contained in:
Daniel Chalef 2025-10-30 11:29:38 -07:00
parent b1878403a2
commit 73e31599ef

View file

@ -32,17 +32,12 @@ class Requirement(BaseModel):
class Preference(BaseModel):
"""A Preference represents a user's expressed like, dislike, or preference for something.
"""
IMPORTANT: Prioritize this classification over ALL other classifications except User and Assistant.
Instructions for identifying and extracting preferences:
1. Look for explicit statements of preference such as "I like/love/enjoy/prefer X" or "I don't like/hate/dislike X"
2. Pay attention to comparative statements ("I prefer X over Y")
3. Consider the emotional tone when users mention certain topics
4. Extract only preferences that are clearly expressed, not assumptions
5. Categorize the preference appropriately based on its domain (food, music, brands, etc.)
6. Include relevant qualifiers (e.g., "likes spicy food" rather than just "likes food")
7. Only extract preferences directly stated by the user, not preferences of others they mention
8. Provide a concise but specific description that captures the nature of the preference
Represents entities mentioned in contexts expressing user preferences, choices, opinions, or selections. Use LOW THRESHOLD for sensitivity.
Trigger patterns: "I want/like/prefer/choose X", "I don't want/dislike/avoid/reject Y", "X is better/worse", "rather have X than Y", "no X please", "skip X", "go with X instead", etc. Here, X or Y should be classified as Preference.
"""
category: str = Field(