Update alz_companion/prompts.py

alz_companion/prompts.py

@@ -282,7 +282,7 @@ Your response MUST be based ONLY on the information in the 'General Guidance' an
 # --- MODIFICATION START ---
 4.  Follow this explicit three-part structure for the paragraph:
     A. **Validate:** Start with a varied, empathetic opening that validates the user’s feeling or concern.
-    B. **Inform with Example:** Next, provide helpful context by briefly referencing a specific, illustrative example from the 'General Guidance' section to show the user this is a known situation (e.g., "This is a common experience; in one story, a character named Anthony often hid his watch when he felt confused.").
+    B. **Inform & Advise from Example:** Next, you MUST find a specific, illustrative example from the 'General Guidance' section. First, briefly state the example (e.g., "This is a common experience; in one story, a character named Anthony..."). Then, you MUST extract the specific, practical advice or strategy described in that same example and present it as the primary recommendation.
     C. **Advise:** Finally, offer 1-2 clear, practical steps from the context that the caregiver can take. If possible, weave in details from 'Relevant Personal Memories' to personalize the advice.
 # --- END OF MODIFICATION ---
 5.  **CRITICAL:** Do not include any preambles, headings, or labels like "My response is...". Address the user directly and naturally.
@@ -376,42 +376,32 @@ User Message: "{question}"
 - Do not add any special formatting or headings.
 """
 
-# In prompts.py, find and REPLACE ANSWER_TEMPLATE_SUMMARIZE
 
+# In prompts.py, replace the ANSWER_TEMPLATE_SUMMARIZE with this stricter version:
 ANSWER_TEMPLATE_SUMMARIZE = """Source excerpts:
 {context}
 
----
-<PARTICIPANTS>
-- Patient's Name: {patient_name}
-- Caregiver's Name: {caregiver_name}
-- Your Role: You are creating this summary for the {role}.
-</PARTICIPANTS>
 ---
 User's Request: {question}
 
 ---
 INSTRUCTIONS FOR THE AI:
-# --- MODIFICATION START ---
---- CRITICAL INSTRUCTIONS ---
-1.  Your ONLY task is to summarize the 'Source excerpts' to directly answer the 'User's Request'.
-2.  **You MUST NOT add any facts or details that are not in the 'Source excerpts'.**
-3.  **CRITICAL:** If the user asks for a summary of a specific time period (e.g., "yesterday", "last three days") or a specific topic (e.g., "daily routine", "word-finding difficulties"), you MUST filter the context and summarize ONLY the information relevant to that specific request.
-4.  **DO NOT provide a general character summary of the person's entire life unless that is what was explicitly asked for.**
----
-# --- END OF MODIFICATION ---
+--- CRITICAL COMMAND ---
+Your ONLY task is to summarize the 'Source excerpts' to directly and ONLY answer the 'User's Request'.
+
+1.  **FILTER:** First, you MUST analyze the "User's Request" to identify any specific constraints like a time period (e.g., "yesterday", "last three days") or a topic (e.g., "daily routine", "word-finding difficulties").
+2.  **EXECUTE:** You MUST then filter the "Source excerpts" and use ONLY the information that directly matches these constraints to build your summary.
+3.  **HALLUCINATION GUARD:** You are forbidden from including any information not present in the filtered source excerpts. DO NOT provide a general summary of the person's entire life unless explicitly asked.
 
 **Formatting and Content Rules:**
-1.  Your final summary MUST be in {language}.
-2.  Write a succinct summary that is tailored to the user's request, addressing the user by their role. For example, "Of course, {caregiver_name}, here is a summary about {patient_name}..."
-3.  The summary should be 5–7 sentences or up to ~120 words.
-4.  Preserve key people, places, and timelines from the source excerpts.
-5.  If the user's request asks for bullet points, use them; otherwise, use a paragraph.
-6.  If the context is empty or irrelevant to the request, state that you could not find information to summarize.
-7.  **CRITICAL:** Do not add any headings, introductions, or concluding remarks. Output only the summary itself.
+- Your final summary MUST be in {language}.
+- Address the user by their role (e.g., "{patient_name}" or "{caregiver_name}").
+- If the context contains no information relevant to the specific request, you MUST state that you could not find any relevant memories.
+- Output ONLY the summary itself.
 """
 
 
+
 # --- PRESERVED: The working multi-hop prompt from prompts_work.py ---
 ANSWER_TEMPLATE_FACTUAL_MULTI = """Context from various sources:
 {context}