Updated

main.py

@@ -5,8 +5,9 @@ from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 from spitch import Spitch
 from langchain.prompts import PromptTemplate
-from langchain_huggingface import HuggingFaceEndpoint   # ✅ updated import
+from langchain_huggingface import HuggingFaceEndpoint
 from langdetect import detect, DetectorFactory
+from huggingface_hub.utils import HfHubHTTPError  # ✅ for quota error handling
 
 DetectorFactory.seed = 0
 
@@ -23,7 +24,7 @@ if not SPITCH_API_KEY:
 os.environ["SPITCH_API_KEY"] = SPITCH_API_KEY
 spitch_client = Spitch()
 
-# ✅ Use new HuggingFaceEndpoint instead of deprecated HuggingFaceHub
+# ✅ HuggingFaceEndpoint
 llm = HuggingFaceEndpoint(
     repo_id=HF_MODEL,
     temperature=0.2,
@@ -57,7 +58,6 @@ Question: {question}
 Answer:
 """
 
-
 stt_chat_template = """You are DevAssist, an AI coding assistant.
 - The input is transcribed speech. Interpret it as a dev question.
 - Provide clear answers with code examples (use markdown triple backticks).
@@ -126,8 +126,15 @@ def root():
 @app.post("/chat")
 def chat(req: ChatRequest, authorization: str | None = Header(None)):
     check_auth(authorization)
-    answer = chat_chain.invoke({"question": req.question})
-    return {"reply": answer.strip() if isinstance(answer, str) else str(answer)}
+    try:
+        answer = chat_chain.invoke({"question": req.question})
+        return {"reply": answer.strip() if isinstance(answer, str) else str(answer)}
+    except HfHubHTTPError as e:
+        if "exceeded" in str(e).lower() or "quota" in str(e).lower():
+            return {
+                "reply": "⚠️ Token quota for today has been used. Please come back in 24 hours."
+            }
+        raise e
 
 @app.post("/stt")
 async def stt_audio(file: UploadFile = File(...), lang_hint: str | None = None, authorization: str | None = Header(None)):
@@ -189,14 +196,21 @@ async def sme_generate(payload: dict = Body(...)):
       "language": "en"
     }
     """
-    response = sme_chain.invoke({
-        "site_name": payload.get("site_name"),
-        "stack": payload.get("stack"),
-        "pages": payload.get("pages"),
-        "language": payload.get("language"),
-    })
-
-    return {"success": True, "data": response}
+    try:
+        response = sme_chain.invoke({
+            "site_name": payload.get("site_name"),
+            "stack": payload.get("stack"),
+            "pages": payload.get("pages"),
+            "language": payload.get("language"),
+        })
+        return {"success": True, "data": response}
+    except HfHubHTTPError as e:
+        if "exceeded" in str(e).lower() or "quota" in str(e).lower():
+            return {
+                "success": False,
+                "error": "⚠️ Token quota for today has been used. Please come back in 24 hours."
+            }
+        raise e
 
 # ✅ Hugging Face requires port 7860, not 8000
 if __name__ == "__main__":