Fix dashboard 404 by serving HTML at root

main.py

@@ -1,69 +1,43 @@
-from fastapi import FastAPI, UploadFile, File
-from fastapi.middleware.cors import CORSMiddleware
+import os
+from fastapi import FastAPI, UploadFile, File, HTTPException
+from fastapi.responses import HTMLResponse, JSONResponse
+from fastapi.staticfiles import StaticFiles
 from agents.visual_analyst import VisualAnalyst
-from agents.memory_agent import MemoryAgent
-from agents.writer_agent import WriterAgent
-
+from dotenv import load_dotenv
+# Load environment variables
+load_dotenv()
 app = FastAPI()
-
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],  # Allows all origins
-    allow_credentials=True,
-    allow_methods=["*"],  # Allows all methods
-    allow_headers=["*"],  # Allows all headers
-)
-
-# Initialize Agents
+# Initialize Agent
 visual_agent = VisualAnalyst()
-memory_agent = MemoryAgent()
-memory_agent.seed_database()
-writer_agent = WriterAgent()
-
-@app.get("/")
-def read_root():
-    return {"message": "MerchFlow AI API is running!"}
-
-@app.post("/analyze-image")
-async def analyze_image(file: UploadFile = File(...)):
-    # Use global agent
-    image_bytes = await file.read()
-    analysis = visual_agent.analyze_image(image_bytes)
-    return analysis
-
-@app.post("/generate-catalog")
-async def generate_catalog(file: UploadFile = File(...)):
+# 1. READ THE DASHBOARD HTML FILE INTO MEMORY
+try:
+    with open("dashboard.html", "r") as f:
+        dashboard_html = f.read()
+except FileNotFoundError:
+    dashboard_html = "<h1>Error: dashboard.html not found. Please ensure the file exists.</h1>"
+# 2. SERVE DASHBOARD AT ROOT (Home Page)
+@app.get("/", response_class=HTMLResponse)
+async def read_root():
+    return dashboard_html
+# 3. KEEP /dashboard ROUTE AS BACKUP
+@app.get("/dashboard", response_class=HTMLResponse)
+async def read_dashboard():
+    return dashboard_html
+@app.post("/analyze")
+async def analyze_merch(file: UploadFile = File(...)):
     try:
-        # Step 1: Visual Analysis (Eyes)
-        image_bytes = await file.read()
-        visual_data = visual_agent.analyze_image(image_bytes)
+        os.makedirs("uploads", exist_ok=True)
+        file_path = f"uploads/{file.filename}"
+        with open(file_path, "wb") as f:
+            f.write(await file.read())
+        result = await visual_agent.analyze_image(file_path)
         
-        if "error" in visual_data:
-             return {"error": "Visual Analysis Failed", "details": visual_data}
-
-        # Step 2: Memory Retrieval (Memory)
-        # Construct query from visual attributes
-        query_parts = []
-        target_keys = ["Main Color", "color", "Material/Texture", "material", "Style/Vibe", "vibe"]
-        for key in target_keys:
-            if key in visual_data:
-                query_parts.append(str(visual_data[key]))
-        
-        search_query = " ".join(query_parts)
-        # Fallback: if no specific keys found, use all string values
-        if not search_query:
-             search_query = " ".join([str(v) for v in visual_data.values() if isinstance(v, str)])
-        
-        seo_keywords = memory_agent.retrieve_keywords(search_query)
-
-        # Step 3: Copywriting (Brain)
-        listing = writer_agent.write_listing(visual_data, seo_keywords)
-
-        return {
-            "visual_data": visual_data,
-            "seo_keywords": seo_keywords,
-            "listing": listing
-        }
+        if os.path.exists(file_path):
+            os.remove(file_path)
+            
+        return JSONResponse(content=result)
     except Exception as e:
-        return {"error": f"Orchestration Error: {str(e)}"}
-
+        return JSONResponse(content={"error": str(e)}, status_code=500)
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=7860)