feat(sofiia-console): expose /metrics and add basic ops counters

Expose Prometheus-style metrics endpoint and add counters for send requests, idempotency replays, and cursor pagination calls, including a safe in-process fallback exposition when prometheus_client is unavailable.

Made-with: Cursor

services/sofiia-console/app/main.py

@@ -57,6 +57,12 @@ from .monitor import collect_all_nodes
 from .ops import run_ops_action, OPS_ACTIONS
 from .docs_router import docs_router
 from . import db as _app_db
+from .metrics import (
+    SOFIIA_SEND_REQUESTS_TOTAL,
+    SOFIIA_IDEMPOTENCY_REPLAYS_TOTAL,
+    SOFIIA_CURSOR_REQUESTS_TOTAL,
+    render_metrics,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -3134,6 +3140,7 @@ async def api_chats_list(
     cursor: Optional[str] = Query(None),
     _auth: str = Depends(require_auth),
 ):
+    SOFIIA_CURSOR_REQUESTS_TOTAL.labels(resource="chats").inc()
     await _ensure_chat_project()
     node_filter = {n.strip().upper() for n in nodes.split(",") if n.strip()}
     cur = _cursor_decode(cursor)
@@ -3236,6 +3243,7 @@ async def api_chat_messages(
     cursor: Optional[str] = Query(None),
     _auth: str = Depends(require_auth),
 ):
+    SOFIIA_CURSOR_REQUESTS_TOTAL.labels(resource="messages").inc()
     cur = _cursor_decode(cursor)
     before_ts = str(cur.get("ts") or "").strip() or None
     before_message_id = str(cur.get("message_id") or "").strip() or None
@@ -3295,6 +3303,7 @@ async def api_chat_send_v2(chat_id: str, body: ChatMessageSendBody, request: Req
     if idem_key:
         cached = _idem_get(chat_id, idem_key)
         if cached:
+            SOFIIA_IDEMPOTENCY_REPLAYS_TOTAL.inc()
             replay = dict(cached)
             replay["idempotency"] = {"replayed": True, "key": idem_key}
             return replay
@@ -3303,6 +3312,7 @@ async def api_chat_send_v2(chat_id: str, body: ChatMessageSendBody, request: Req
     info = _parse_chat_id(chat_id)
     target_node = ((body.routing or {}).get("force_node_id") or info["node_id"] or "NODA2").upper()
     target_agent = info["agent_id"] or "sofiia"
+    SOFIIA_SEND_REQUESTS_TOTAL.labels(node_id=target_node).inc()
     project_id = body.project_id or CHAT_PROJECT_ID
     session_id = body.session_id or chat_id
     user_id = body.user_id or "console_user"
@@ -3377,6 +3387,12 @@ async def api_chat_send_v2(chat_id: str, body: ChatMessageSendBody, request: Req
     return result
 
 
+@app.get("/metrics")
+def metrics():
+    data, content_type = render_metrics()
+    return Response(content=data, media_type=content_type)
+
+
 @app.post("/api/chat/send")
 async def api_chat_send(body: ChatSendBody, request: Request):
     """BFF chat: Ollama or router. Returns runtime contract fields. Rate: 30/min."""