cron manager fix and documents indentation fixes (#509)

CONTRIBUTING.md

@@ -212,10 +212,12 @@ All FastAPI route handlers use Pydantic models for request/response validation:
 from pydantic import BaseModel
 from fastapi import FastAPI
 
+
 class UserRequest(BaseModel):
     name: str
     age: int
 
+
 @app.post("/users")
 async def create_user(user: UserRequest):
     # Pydantic validates name is string, age is int

DSL/CronManager/script/store_secrets_in_vault.sh

@@ -165,7 +165,14 @@ setup_python_environment() {
         echo "[$(date '+%Y-%m-%d %H:%M:%S')] ERROR: Failed to install loguru" >&2
         return 1
     }
-
+
+    # Install requests, required by LokiLogger which decrypt_vault_secrets.py imports
+    echo "[$(date '+%Y-%m-%d %H:%M:%S')] Installing requests library..."
+    "$UV_BIN" pip install --python "$VENV_PATH/bin/python3" "requests>=2.32" 2>&1 || {
+        echo "[$(date '+%Y-%m-%d %H:%M:%S')] ERROR: Failed to install requests" >&2
+        return 1
+    }
+
     # Mark setup as complete
     touch "$VENV_PATH/.setup_complete"
 
@@ -196,6 +203,9 @@ if ! setup_python_environment; then
     exit 1
 fi
 
+# Set Python path
+export PYTHONPATH="/app:/app/src:/app/src/vector_indexer:$PYTHONPATH"
+
 # Function to determine platform name
 get_platform_name() {
     case "$llmPlatform" in

docs/API_TOOL_CALLING.md

@@ -209,21 +209,21 @@ A `PointStruct` is built:
 
 ```python
 PointStruct(
-    id      = endpoint_id,            # UUID used directly as Qdrant point ID
-    vector  = {
-        "dense":  [v1, v2, ..., v3072],
+    id=endpoint_id,  # UUID used directly as Qdrant point ID
+    vector={
+        "dense": [v1, v2, ..., v3072],
         "sparse": {"indices": [...], "values": [...]},
     },
-    payload = {                        # Stored metadata — no extra DB lookup needed
-        "endpoint_id":      "...",
-        "name":             "get_national_holidays",
-        "description":      "...",
-        "url":              "https://openholidaysapi.org/PublicHolidays",
-        "method":           "GET",
-        "params":           [...],
+    payload={  # Stored metadata — no extra DB lookup needed
+        "endpoint_id": "...",
+        "name": "get_national_holidays",
+        "description": "...",
+        "url": "https://openholidaysapi.org/PublicHolidays",
+        "method": "GET",
+        "params": [...],
         "enriched_context": "...",
-        "service_id":       "...",
-    }
+        "service_id": "...",
+    },
 )
 ```
 
@@ -306,9 +306,9 @@ Instantiated once in `ToolClassifier.__init__()` and reuses the shared Qdrant `h
 
 ```python
 APISemanticSearcher(
-    embedding_service=orchestration_service,   # generates dense embeddings
-    qdrant_client=self._qdrant_client,         # shared connection pool
-    disambiguator=None,                        # optional: inject for testing
+    embedding_service=orchestration_service,  # generates dense embeddings
+    qdrant_client=self._qdrant_client,  # shared connection pool
+    disambiguator=None,  # optional: inject for testing
 )
 ```
 
@@ -861,7 +861,7 @@ Results are **deduplicated by endpoint name** (a single endpoint matched by two
 
 ```python
 class ExecutionMode(str, Enum):
-    SINGLE   = "single"
+    SINGLE = "single"
     PARALLEL = "parallel"
 ```
 
@@ -947,7 +947,7 @@ await multi_loop.stream_run_turn(
     chat_id=chat_id,
     user_message=request.message,
     conversation_history=conversation_history,
-    endpoint_states=session.parallel_endpoints,   # list[EndpointSessionState]
+    endpoint_states=session.parallel_endpoints,  # list[EndpointSessionState]
     turn_count=session.turn_count,
     awaiting_continuation=session.awaiting_continuation,
     session_language=effective_session_language,
@@ -979,7 +979,9 @@ call_payloads = [
     {**state.endpoint, "call_params": state.collected_params}
     for state in parallel_endpoints
 ]
-multi_result = await MultiAPICaller(api_caller).call_all(call_payloads, language=detected_language)
+multi_result = await MultiAPICaller(api_caller).call_all(
+    call_payloads, language=detected_language
+)
 ```
 
 ---
@@ -1402,7 +1404,7 @@ endpoint A sends a message that strongly matches endpoint B), both the session a
 L2 key are cleaned up:
 
 ```python
-await session_store.delete(request.chatId)        # existing behaviour
+await session_store.delete(request.chatId)  # existing behaviour
 if FeatureFlags.ATC_RESPONSE_CACHE_ENABLED:
     await ATCCacheStore().invalidate_l2(request.chatId)
 ```

docs/CONTEXT_WORKFLOW_GREETING_DETECTION.md

@@ -223,17 +223,17 @@ In **Phase 2**, `ContextWorkflowExecutor` calls `get_greeting_response(greeting_
 
 ```python
 GREETINGS_ET = {
-    "hello":   "Tere! Kuidas ma saan sind aidata?",
+    "hello": "Tere! Kuidas ma saan sind aidata?",
     "goodbye": "Nägemist! Head päeva!",
-    "thanks":  "Palun! Kui on veel küsimusi, küsi julgelt.",
-    "casual":  "Tere! Mida ma saan sinu jaoks teha?",
+    "thanks": "Palun! Kui on veel küsimusi, küsi julgelt.",
+    "casual": "Tere! Mida ma saan sinu jaoks teha?",
 }
 
 GREETINGS_EN = {
-    "hello":   "Hello! How can I help you?",
+    "hello": "Hello! How can I help you?",
     "goodbye": "Goodbye! Have a great day!",
-    "thanks":  "You're welcome! Feel free to ask if you have more questions.",
-    "casual":  "Hey! What can I do for you?",
+    "thanks": "You're welcome! Feel free to ask if you have more questions.",
+    "casual": "Hey! What can I do for you?",
 }
 ```
 
@@ -358,29 +358,31 @@ Key log entries emitted during a request:
 
 ```python
 class ConversationRound(BaseModel):
-    user_message: str   # The user's message text
-    bot_message: str    # The bot's response text
-    timestamp: float    # Unix timestamp of the round
+    user_message: str  # The user's message text
+    bot_message: str  # The bot's response text
+    timestamp: float  # Unix timestamp of the round
 ```
 
 ### `ConversationHistoryState` (Redis fetch result)
 
 ```python
 class ConversationHistoryState(BaseModel):
-    chat_id: str                        # Unique conversation identifier
-    rounds: list[ConversationRound]     # Ordered rounds (newest last), capped at 10
-    summary: Optional[str]              # Incremental summary of evicted older rounds
+    chat_id: str  # Unique conversation identifier
+    rounds: list[ConversationRound]  # Ordered rounds (newest last), capped at 10
+    summary: Optional[str]  # Incremental summary of evicted older rounds
 ```
 
 ### `ContextDetectionResult` (Phase 1 output)
 
 ```python
 class ContextDetectionResult(BaseModel):
-    is_greeting: bool               # True if query is a greeting
-    greeting_type: str              # "hello" | "goodbye" | "thanks" | "casual"
-    can_answer_from_context: bool   # True if query can be answered from history or summary
-    reasoning: str                  # LLM's brief explanation
-    answered_from_summary: bool     # True when answer derived from summary path
+    is_greeting: bool  # True if query is a greeting
+    greeting_type: str  # "hello" | "goodbye" | "thanks" | "casual"
+    can_answer_from_context: (
+        bool  # True if query can be answered from history or summary
+    )
+    reasoning: str  # LLM's brief explanation
+    answered_from_summary: bool  # True when answer derived from summary path
     context_snippet: Optional[str]  # Relevant excerpt for Phase 2 generation, or None
 ```
 

docs/HYBRID_SEARCH_CLASSIFICATION.md

@@ -115,9 +115,7 @@ tokens = re.findall(r"\w+", text.lower())  # ["mis", "suhe", "on", "euro", ...]
 
 ```python
 # Collection: "intent_collections"
-vectors_config = {
-    "dense": VectorParams(size=3072, distance=Distance.COSINE)
-}
+vectors_config = {"dense": VectorParams(size=3072, distance=Distance.COSINE)}
 sparse_vectors_config = {
     "sparse": SparseVectorParams(index=SparseIndexParams(on_disk=False))
 }
@@ -197,12 +195,12 @@ Queries Qdrant using only the dense vector to get **actual cosine similarity sco
 
 ```python
 # classifier.py → _dense_search()
-POST /collections/intent_collections/points/query
+POST / collections / intent_collections / points / query
 {
     "query": [0.023, -0.041, ...],  # 3072-dim dense vector
     "using": "dense",
-    "limit": 6,                     # DENSE_SEARCH_TOP_K * 2 (3 * 2 = 6, allows dedup)
-    "with_payload": true
+    "limit": 6,  # DENSE_SEARCH_TOP_K * 2 (3 * 2 = 6, allows dedup)
+    "with_payload": true,
 }
 ```
 
@@ -219,15 +217,15 @@ Sparse prefetch is only included if the query produces a non-empty sparse vector
 ```python
 # classifier.py → _hybrid_search()
 # First checks collection exists and has data (points_count > 0)
-POST /collections/intent_collections/points/query
+POST / collections / intent_collections / points / query
 {
     "prefetch": [
         {"query": dense_vector, "using": "dense", "limit": 10},
-        {"query": {"indices": [...], "values": [...]}, "using": "sparse", "limit": 10}
+        {"query": {"indices": [...], "values": [...]}, "using": "sparse", "limit": 10},
     ],
     "query": {"fusion": "rrf"},
     "limit": 5,
-    "with_payload": true
+    "with_payload": true,
 }
 ```
 

docs/REDIS_SESSION_STORE.md

@@ -65,9 +65,9 @@ if session is None:
     # No active session → this is a fresh conversation
     ...
 else:
-    print(session.state)            # "collecting_params"
-    print(session.collected_params) # {"city": "Tallinn"}
-    print(session.turn_count)       # 2
+    print(session.state)  # "collecting_params"
+    print(session.collected_params)  # {"city": "Tallinn"}
+    print(session.turn_count)  # 2
 ```
 
 ---
@@ -129,12 +129,14 @@ session_store = request.app.state.session_store
 session = await session_store.get(request.chatId)
 if session is None:
     detected_endpoint = ...  # endpoint detected from user query
-    await session_store.save(APIToolSession(
-        chat_id=request.chatId,
-        state="collecting_params",
-        selected_endpoint=detected_endpoint,
-        turn_count=1,
-    ))
+    await session_store.save(
+        APIToolSession(
+            chat_id=request.chatId,
+            state="collecting_params",
+            selected_endpoint=detected_endpoint,
+            turn_count=1,
+        )
+    )
     return "Which city would you like weather for?"
 
 # --- Turn 2+ ---

docs/TESTPRODUCTIONLLM_SERVICE_WORKFLOW.md

@@ -212,13 +212,15 @@ orchestration_service = self.orchestration_service
 service_content = service_result["content"]
 service_buttons = service_result["buttons"]
 
+
 async def service_stream() -> AsyncIterator[str]:
     yield orchestration_service.format_sse(
         chat_id, service_content, service_buttons or None
     )
     yield orchestration_service.format_sse(chat_id, "END")
     orchestration_service.log_costs(costs_metric)
 
+
 return service_stream()
 ```
 
@@ -229,12 +231,13 @@ return service_stream()
 **`llm_orchestration_service.py` → `format_sse()`** (line 1195):
 
 ```python
-def format_sse(self, chat_id: str, content: str,
-               buttons: Optional[List[Dict[str, Any]]] = None) -> str:
+def format_sse(
+    self, chat_id: str, content: str, buttons: Optional[List[Dict[str, Any]]] = None
+) -> str:
     inner_payload: Dict[str, Any] = {"content": content}
     if buttons:
         inner_payload["buttons"] = buttons
-    
+
     payload = {
         "chatId": chat_id,
         "payload": inner_payload,