Update app.py

app.py

@@ -1,4 +1,5 @@
 import os
+import time
 from collections import deque
 from flask import Flask, request, jsonify
 from waitress import serve
@@ -24,8 +25,15 @@ client = genai.Client(api_key=os.environ.get("GEMINI_API_KEY"))
 MAX_MESSAGES = 20  # user+assistant messages combined
 HISTORY = deque(maxlen=MAX_MESSAGES)  # holds types.Content objects
 
+
+def _client_ip() -> str:
+    # HF may proxy requests; this is best-effort
+    return request.headers.get("x-forwarded-for", request.remote_addr or "unknown")
+
+
 @app.get("/health")
 def health():
+    print(f"[/health] {time.strftime('%Y-%m-%d %H:%M:%S')} ip={_client_ip()} mem={len(HISTORY)}/{MAX_MESSAGES}")
     return jsonify({
         "ok": True,
         "model": MODEL,
@@ -34,13 +42,23 @@ def health():
         "max_messages": MAX_MESSAGES,
     })
 
+
 @app.post("/v1/chat")
 def chat():
+    t0 = time.time()
+    ip = _client_ip()
+
     data = request.get_json(silent=True) or {}
     user_text = (data.get("text") or "").strip()
+
+    print(f"[/v1/chat] START {time.strftime('%Y-%m-%d %H:%M:%S')} ip={ip} mem_before={len(HISTORY)}/{MAX_MESSAGES}")
+
     if not user_text:
+        print(f"[/v1/chat] ERROR missing text ip={ip}")
         return jsonify({"error": "Missing 'text'"}), 400
 
+    print(f"[/v1/chat] user_text_len={len(user_text)} user_text={user_text!r}")
+
     # Add user message to memory
     HISTORY.append(
         types.Content(
@@ -48,6 +66,7 @@ def chat():
             parts=[types.Part.from_text(text=user_text)],
         )
     )
+    print(f"[/v1/chat] appended user msg -> mem_now={len(HISTORY)}/{MAX_MESSAGES}")
 
     config = types.GenerateContentConfig(
         system_instruction=[types.Part.from_text(text=SYSTEM_PROMPT)],
@@ -61,6 +80,7 @@ def chat():
             config=config,
         )
         reply_text = (resp.text or "").strip()
+        print(f"[/v1/chat] gemini_ok reply_len={len(reply_text)}")
 
         # Add assistant message to memory
         HISTORY.append(
@@ -69,12 +89,17 @@ def chat():
                 parts=[types.Part.from_text(text=reply_text)],
             )
         )
+        print(f"[/v1/chat] appended model msg -> mem_now={len(HISTORY)}/{MAX_MESSAGES}")
+
+        dt_ms = int((time.time() - t0) * 1000)
+        print(f"[/v1/chat] END ip={ip} total_ms={dt_ms}")
 
         return jsonify({
             "input": user_text,
             "reply_text": reply_text,
             "model": MODEL,
             "memory_messages": len(HISTORY),
+            "total_ms": dt_ms,
         })
 
     except Exception as e:
@@ -83,14 +108,23 @@ def chat():
         # Roll back last user message on failure
         if len(HISTORY) > 0 and getattr(HISTORY[-1], "role", None) == "user":
             HISTORY.pop()
+            print(f"[/v1/chat] rollback user msg -> mem_now={len(HISTORY)}/{MAX_MESSAGES}")
+
+        dt_ms = int((time.time() - t0) * 1000)
+        print(f"[/v1/chat] FAIL ip={ip} total_ms={dt_ms}")
 
         return jsonify({"error": "Gemini call failed"}), 500
 
+
 @app.post("/v1/reset")
 def reset():
+    ip = _client_ip()
+    print(f"[/v1/reset] {time.strftime('%Y-%m-%d %H:%M:%S')} ip={ip} clearing mem (was {len(HISTORY)}/{MAX_MESSAGES})")
     HISTORY.clear()
     return jsonify({"ok": True, "memory_messages": 0})
 
+
 if __name__ == "__main__":
     port = int(os.environ.get("PORT", "7860"))
+    print(f"[startup] model={MODEL} thinking_level={THINKING_LEVEL} max_messages={MAX_MESSAGES} port={port}")
     serve(app, host="0.0.0.0", port=port)