diff --git a/src/pydantic_censoring.py b/src/pydantic_censoring.py
index c0145ba..4323c5a 100644
--- a/src/pydantic_censoring.py
+++ b/src/pydantic_censoring.py
@@ -1,6 +1,6 @@
 import asyncio
-from typing import Set, List, Tuple
-from pydantic import BaseModel
+from typing import List, Dict, Annotated
+from pydantic import BaseModel, Field
 from pydantic_ai import Agent
 from pydantic_ai.models.openai import OpenAIChatModel, Model
 from pydantic_ai.providers.openai import OpenAIProvider
@@ -27,7 +27,9 @@ messages_to_censor = [
 
 class SensitiveData(BaseModel):
     """Structure for identifying sensitive data that should be censored"""
-    sensitive_items: Set[str]
+    sensitive_items: List[str] = Field(
+        description="A list of sensitive data items (names, IDs, emails, etc.) that should be censored from the text"
+    )
 
 
 async def simple_test_response(model: Model):
@@ -45,7 +47,7 @@ async def simple_test_response(model: Model):
         print()  # Add a final newline
 
 
-async def get_words_to_censor(model: Model, prompt: str) -> Set[str]:
+async def get_words_to_censor(model: Model, prompt: str) -> List[str]:
     # Make a structured response that will return everything from the prompt that needs censoring.
 
     # Create an agent that returns structured data
@@ -56,16 +58,24 @@ async def get_words_to_censor(model: Model, prompt: str) -> Set[str]:
     )
 
     result = await censor_agent.run(prompt)
-    sensitive_items = set(sorted(list(result.output.sensitive_items)))  # for sorting
+    sensitive_items = sorted(result.output.sensitive_items)  # for sorting
     return sensitive_items
 
 
+SensitiveTextList = Annotated[
+    List[str],
+    Field(description="List of original sensitive text values that should be replaced with this placeholder")
+]
+
+
 class PlaceholderMapping(BaseModel):
     """Structure for mapping sensitive data to placeholders"""
-    mappings: List[Tuple[str, List[str]]]  # List of (placeholder, [original_texts]) tuples
+    mappings: Dict[str, SensitiveTextList] = Field(
+        description="Dictionary where keys are placeholder names (e.g., '[PERSON_1]', '[EMAIL_1]') and values are lists of original sensitive text items"
+    )
 
 
-async def generate_placeholders(model: Model, censored_words: Set[str]) -> List[Tuple[str, List[str]]]:
+async def generate_placeholders(model: Model, censored_words: List[str]) -> Dict[str, List[str]]:
     """Generate placeholders for censored words"""
     placeholder_agent = Agent(
         model,
@@ -74,8 +84,7 @@ async def generate_placeholders(model: Model, censored_words: Set[str]) -> List[
     )
 
     # Convert set to sorted list for consistent ordering
-    words_list = sorted(list(censored_words))
-    result = await placeholder_agent.run(words_list)
+    result = await placeholder_agent.run(censored_words)
     return result.output.mappings
 
 
@@ -90,9 +99,9 @@ async def main():
         )
     )
 
-    censored_words: Set = set()
+    censored_words: List[str] = []
     for message in messages_to_censor:
-        censored_words = censored_words | await get_words_to_censor(model, message)
+        censored_words += await get_words_to_censor(model, message)
 
     print("\nWords to censor:")
     for word in censored_words:
@@ -103,8 +112,14 @@ async def main():
     placeholder_mappings = await generate_placeholders(model, censored_words)
 
     print("\nPlaceholder mappings:")
-    for placeholder, originals in placeholder_mappings:
-        print(f"  {placeholder} → {', '.join(originals)}")
+    for placeholder, sensitive_items in placeholder_mappings.items():
+        print(f"  {placeholder}: {sensitive_items}")
+
+    print(f'Sensitive words: {len(censored_words)}')
+    total_count = 0
+    for _, value_list in placeholder_mappings.items():
+        total_count += len(value_list)
+    print(f'Sensitive words in placeholders: {total_count}')
 
 
 if __name__ == '__main__':