started to add use cases Json_Prompt__Create_Summary and Json_Prompt__Validate_Summary

Author: DinisCruz

osbot_llms/llms/prompt_to_json/use_cases/Json_Prompt__Create_Summary.py

@@ -0,0 +1,49 @@
+from pydantic import BaseModel
+
+from osbot_llms.llms.prompt_to_json.Prompt_To_Json__Open_AI import Prompt_To_Json__Open_AI
+from osbot_utils.base_classes.Type_Safe import Type_Safe
+
+
+class Model__Response_Format__Json_Prompt__Create_Summary(BaseModel):
+    summary: str
+    keywords: list[str]
+
+class Json_Prompt__Create_Summary(Type_Safe):
+    prompt_to_json  : Prompt_To_Json__Open_AI
+    response_format : type                      = None
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.response_format = Model__Response_Format__Json_Prompt__Create_Summary
+
+
+    def create_summary(self, target_text: str):
+        with self.prompt_to_json as _:
+            _.set_model__gpt_4o_mini (                    )
+            _.set_response_format    (self.response_format)
+            _.add_message__system    (self.system_prompt())
+            _.add_message__user      (target_text         )
+
+            return _.invoke()
+
+    def system_prompt(self):
+        return """
+You are a precise text summarization system. Follow these rules exactly:
+
+1. ONLY use information explicitly stated in the input text
+2. DO NOT add any external knowledge, context, or assumptions
+3. DO NOT include any dates, statistics, or specific details unless they appear verbatim in the text
+4. DO NOT make generalizations about the topic
+5. DO NOT add explanatory or background information
+6. For keywords:
+   - ONLY extract words/phrases that appear in the original text
+   - List them in order of appearance
+   - DO NOT add related terms or categories
+   - Limit to 5 keywords maximum
+
+FORMAT:
+- Summary: One or two sentences using only information from the text
+- Keywords: Maximum 5 terms that appear verbatim in the text
+
+If the input text is too short or lacks sufficient content, just provide a very short summary:
+"""

osbot_llms/llms/prompt_to_json/use_cases/Json_Prompt__Validate_Summary.py

@@ -0,0 +1,44 @@
+from pydantic import BaseModel
+from osbot_llms.llms.prompt_to_json.Prompt_To_Json__Open_AI import Prompt_To_Json__Open_AI
+from osbot_utils.base_classes.Type_Safe import Type_Safe
+
+class Model__Response_Format__Json_Prompt__Validate_Summary(BaseModel):
+    is_valid     : bool
+    confidence   : float
+    quality      : float
+    issues_found : list[str]
+
+class Json_Prompt__Validate_Summary(Type_Safe):
+    prompt_to_json  : Prompt_To_Json__Open_AI
+    response_format : type = None
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.response_format = Model__Response_Format__Json_Prompt__Validate_Summary
+
+    def validate_summary(self, original_text: str, summary: str):
+        with self.prompt_to_json as _:
+            _.set_model__gpt_4o_mini()
+            _.set_response_format(self.response_format)
+            _.add_message__system(self.system_prompt())
+            _.add_message__user(f"""Original Text:
+{original_text}
+
+Summary to validate:
+{summary}""")
+            return _.invoke()
+
+    def system_prompt(self):
+        return """Analyze if the provided summary accurately represents the original text.
+Rules:
+1. Check if summary ONLY contains information from original text
+2. Verify no external knowledge/assumptions added
+3. Confirm dates/stats appear in original
+4. Check for over generalizations
+5. Look for added explanatory content
+
+Score explanation:
+- confidence: your certainty in the assessment
+- quality: how well summary reflects original content
+
+Return detailed issues list for any problems found."""