Update app.py

app.py

@@ -27,7 +27,7 @@ client = chromadb.CloudClient(
     tenant=tenant,
     database=database
 )
-print(f"Connection to chromabd successful {database} {api_key}")
+print(f"Connection to chromabd successful...")
         
 # === COLLECTION VALIDATION ===
 # Ensure the required collection exists and has data
@@ -350,42 +350,75 @@ def get_available_sources() -> Dict[str, Any]:
     except Exception as e:
         return {"error": f"Failed to get sources: {str(e)}", "sources": [], "success": False}
 
-
-def slice_list(lst: list, start: int, end: int) -> list:
+# MCP Tool Definitions
+def search_rs_studies(
+    query: str, 
+    num_results: int = 5, 
+    source_filter: Optional[str] = None,
+    task_type: str = "search"
+) -> str:
     """
-    A tool that slices a list given a start and end index.
+    Search the RS Studies knowledge base for relevant information.
+    
+    This tool provides semantic search across RS trading system documentation,
+    Chennai meetup transcripts, and Q&A content with optimized EmbeddingGemma prompts.
+    
     Args:
-        lst: The list to slice.
-        start: The start index.
-        end: The end index.
+        query: Your search question or topic (required)
+        num_results: Number of results to return (1-50, default: 5)
+        source_filter: Limit search to specific source:
+                      - 'rs_stkege_01': RS trading system documentation
+                      - 'cheenai_meet_full': Chennai meetup transcripts  
+                      - 'QnAYoutubeChannel': Q&A discussions
+                      - None: Search all sources (default)
+        task_type: Search optimization using EmbeddingGemma task-specific prompts:
+                  - 'search'/'retrieval_query': General search (default)
+                  - 'question'/'question_answering': Question answering format
+                  - 'fact'/'fact_checking': Fact checking format
+                  - 'classification': Text classification tasks
+                  - 'clustering': Document clustering and grouping
+                  - 'semantic_similarity': Semantic similarity assessment
+                  - 'code_retrieval': Code search and retrieval
+    
     Returns:
-        The sliced list.
+        JSON string with search results including content, sources, and similarity scores
     """
-    return lst[start:end]
+    # Validate parameters
+    if not query or not query.strip():
+        return json.dumps({"error": "Query cannot be empty", "results": [], "success": False})
+    
+    num_results = max(1, min(num_results, config.MAX_NUM_RESULTS))
+    
+    if source_filter and source_filter not in config.VALID_SOURCES:
+        return json.dumps({
+            "error": f"Invalid source_filter. Must be one of: {config.VALID_SOURCES}",
+            "results": [],
+            "success": False
+        })
+    
+    valid_task_types = list(EmbeddingGemmaPrompts.TASKS.keys())
+    if task_type not in valid_task_types:
+        return json.dumps({
+            "error": f"Invalid task_type. Must be one of: {valid_task_types}",
+            "results": [],
+            "success": False
+        })
+    
+    # Perform search
+    results = search_knowledge_base(query, num_results, source_filter, task_type)
+    return json.dumps(results, indent=2)
 
-def say_hello(name: str) -> str:
-    """
-    A tool that returns message "Hello <name>"
-    Args:
-        str: name to say hello to.
-    Returns:
-        Hello <name>
-    """
-    return f"Hello, {name}"
 
 with gr.Blocks() as demo:
     gr.Markdown(
         """
-        This is a demo of a MCP-only tool.
-        This tool slices a list.
+        This is a MCP only tool for RS Studies
+        This connects to a remote chromadb instance.
         This tool is MCP-only, so it does not have a UI.
         """
     )
     gr.api(
-        slice_list
-    )
-    gr.api(
-        say_hello
+        search_rs_studies
     )
 
 _, url, _ = demo.launch(mcp_server=True)