Update app.py

app.py

@@ -1,8 +1,6 @@
 """
-AURA Chat — Gradio Space + MCP Server Tab
-Single-file Gradio app that:
-- Chat / Analysis Tab: Enter prompts, analyze, and chat with the assistant.
-- MCP Server Tab: Call scraping and analysis functions directly with JSON output.
+AURA Chat — Hedge Fund Picks
+Enhanced UI/UX with instructions, examples, YouTube video, and colored containers.
 """
 
 import os
@@ -15,7 +13,7 @@ import traceback
 from typing import List
 import gradio as gr
 
-# Defensive: fresh event loop early to avoid fd race on shutdown
+# ===================== Defensive event loop =====================
 if sys.platform != "win32":
     try:
         loop = asyncio.new_event_loop()
@@ -23,43 +21,26 @@ if sys.platform != "win32":
     except Exception:
         traceback.print_exc()
 
-# =============================================================================
-# CONFIGURATION (fixed)
-# =============================================================================
+# ===================== CONFIGURATION =====================
 SCRAPER_API_URL = os.getenv("SCRAPER_API_URL", "https://deep-scraper-96.created.app/api/deep-scrape")
 SCRAPER_HEADERS = {
     "User-Agent": "Mozilla/5.0",
     "Content-Type": "application/json"
 }
-
 LLM_MODEL = os.getenv("LLM_MODEL", "openai/gpt-oss-20b:free")
 MAX_TOKENS = int(os.getenv("LLM_MAX_TOKENS", "3000"))
 SCRAPE_DELAY = float(os.getenv("SCRAPE_DELAY", "1.0"))
-
 OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
 OPENAI_BASE_URL = os.getenv("OPENAI_BASE_URL", "https://openrouter.ai/api/v1")
 
 PROMPT_TEMPLATE = f"""You are AURA, a concise, professional hedge-fund research assistant.
 Task:
-- Given scraped data below, produce a clear, readable analysis that:
-  1) Lists the top 5 stock picks (or fewer if not enough data).
-  2) For each stock provide: Ticker / Company name, short rationale (2-3 bullets),
-     and an explicit **Investment Duration** entry: a one-line "When to Invest"
-     and a one-line "When to Sell" instruction (these two lines are mandatory
-     for each stock).
-  3) Keep each stock entry short and scannable. Use a bullet list or numbered list.
-  4) At the top, provide a 2-3 sentence summary conclusion (market context +
-     highest conviction pick).
-  5) Output in plain text, clean formatting, easy for humans to read. No JSON.
-  6) After the list, include a concise "Assumptions & Risks" section (2-3 bullet points).
-Important: Be decisive. If data is insufficient, state that clearly and provide
-the best-available picks with lower confidence.
-Max tokens for the LLM response: {MAX_TOKENS}
-Model: {LLM_MODEL}"""
+- Given scraped data below, produce a clear analysis listing top stocks (with Investment Duration)
+- Output must be human-readable text, 2-3 sentence summary, 5 top stocks max, and Assumptions & Risks
+Max tokens: {MAX_TOKENS}, Model: {LLM_MODEL}
+"""
 
-# =============================================================================
-# SCRAPING HELPERS
-# =============================================================================
+# ===================== SCRAPING HELPERS =====================
 def deep_scrape(query: str, retries: int = 3, timeout: int = 40) -> str:
     payload = {"query": query}
     last_err = None
@@ -69,8 +50,7 @@ def deep_scrape(query: str, retries: int = 3, timeout: int = 40) -> str:
             resp.raise_for_status()
             data = resp.json()
             if isinstance(data, dict):
-                parts = [f"{k.upper()}:\n{v}\n" for k, v in data.items()]
-                return "\n".join(parts)
+                return "\n".join([f"{k.upper()}:\n{v}\n" for k, v in data.items()])
             return str(data)
         except Exception as e:
             last_err = e
@@ -85,36 +65,33 @@ def multi_scrape(queries: List[str], delay: float = SCRAPE_DELAY) -> str:
         if not q:
             continue
         aggregated.append(f"\n=== QUERY: {q} ===\n")
-        scraped = deep_scrape(q)
-        aggregated.append(scraped)
+        aggregated.append(deep_scrape(q))
         time.sleep(delay)
     return "\n".join(aggregated)
 
-# =============================================================================
-# LLM INTERACTION
-# =============================================================================
+# ===================== LLM INTERACTION =====================
 try:
     from openai import OpenAI
 except Exception:
     OpenAI = None
 
-def run_llm_system_and_user(system_prompt: str, user_text: str, model: str = LLM_MODEL, max_tokens: int = MAX_TOKENS) -> str:
+def run_llm_system_and_user(system_prompt: str, user_text: str) -> str:
     if OpenAI is None:
-        return "ERROR: openai package not installed or available."
+        return "ERROR: openai package not installed."
     if not OPENAI_API_KEY:
-        return "ERROR: OPENAI_API_KEY not set in environment."
+        return "ERROR: OPENAI_API_KEY not set."
     client = None
     try:
         client = OpenAI(base_url=OPENAI_BASE_URL, api_key=OPENAI_API_KEY)
         completion = client.chat.completions.create(
-            model=model,
+            model=LLM_MODEL,
             messages=[{"role": "system", "content": system_prompt}, {"role": "user", "content": user_text}],
-            max_tokens=max_tokens,
+            max_tokens=MAX_TOKENS,
         )
-        if hasattr(completion, "choices") and len(completion.choices) > 0:
+        if hasattr(completion, "choices") and completion.choices:
             try:
                 return completion.choices[0].message.content
-            except Exception:
+            except:
                 return str(completion.choices[0])
         return str(completion)
     except Exception as e:
@@ -124,164 +101,117 @@ def run_llm_system_and_user(system_prompt: str, user_text: str, model: str = LLM
             if client is not None:
                 try:
                     client.close()
-                except Exception:
-                    try:
-                        asyncio.get_event_loop().run_until_complete(client.aclose())
-                    except Exception:
-                        pass
-        except Exception:
-            pass
+                except:
+                    try: asyncio.get_event_loop().run_until_complete(client.aclose())
+                    except: pass
+        except: pass
 
-# =============================================================================
-# MAIN PIPELINE
-# =============================================================================
+# ===================== MAIN PIPELINE =====================
 def analyze_and_seed_chat(prompts_text: str):
     if not prompts_text.strip():
-        return "Please enter at least one prompt (query).", []
+        return "Please enter at least one prompt.", []
     queries = [line.strip() for line in prompts_text.splitlines() if line.strip()]
-    scraped = multi_scrape(queries, delay=SCRAPE_DELAY)
+    scraped = multi_scrape(queries)
     if scraped.startswith("ERROR"):
         return scraped, []
-    user_payload = f"SCRAPED DATA:\n\n{scraped}\n\nPlease follow the system instructions and output the analysis."
+    user_payload = f"SCRAPED DATA:\n\n{scraped}\n\nGenerate analysis."
     analysis = run_llm_system_and_user(PROMPT_TEMPLATE, user_payload)
     if analysis.startswith("ERROR"):
         return analysis, []
     initial_chat = [
-        {"role": "user", "content": f"Analyze the data I provided (prompts: {', '.join(queries)})"},
+        {"role": "user", "content": f"Analyze prompts: {', '.join(queries)}"},
         {"role": "assistant", "content": analysis}
     ]
     return analysis, initial_chat
 
 def continue_chat(chat_messages, user_message: str, analysis_text: str):
-    if chat_messages is None:
-        chat_messages = []
-    if not user_message.strip():
-        return chat_messages
+    if chat_messages is None: chat_messages = []
+    if not user_message.strip(): return chat_messages
     chat_messages.append({"role": "user", "content": user_message})
-    followup_system = (
-        "You are AURA, a helpful analyst. Use the previous analysis as context; answer follow-ups concisely."
-    )
+    followup_system = "You are AURA. Use previous analysis as context; answer follow-ups concisely."
     user_payload = f"REFERENCE ANALYSIS:\n\n{analysis_text}\n\nUSER QUESTION: {user_message}"
     assistant_reply = run_llm_system_and_user(followup_system, user_payload)
     chat_messages.append({"role": "assistant", "content": assistant_reply})
     return chat_messages
 
-def convert_to_gradio_chat_format(chat_messages):
-    return chat_messages or []
-
-# =============================================================================
-# MCP SERVER HELPERS
-# =============================================================================
-def mcp_scrape(query: str):
-    return {"query": query, "result": deep_scrape(query)}
-
-def mcp_multi_scrape(queries_text: str):
-    queries = [line.strip() for line in queries_text.splitlines() if line.strip()]
-    return {"queries": queries, "result": multi_scrape(queries)}
-
-def mcp_analyze(prompts_text: str):
-    analysis, seed_chat = analyze_and_seed_chat(prompts_text)
-    return {"prompts": prompts_text, "analysis": analysis, "seed_chat": seed_chat}
-
-# =============================================================================
-# GRADIO UI
-# =============================================================================
-def build_demo_with_mcp():
-    with gr.Blocks(title="AURA Chat — Hedge Fund Picks + MCP Server") as demo:
-        with gr.Tabs():
-            # Chat / Analysis Tab
-            with gr.TabItem("Chat / Analysis"):
-                with gr.Row():
-                    with gr.Column(scale=1):
-                        prompts = gr.Textbox(lines=6, label="Data Prompts")
-                        analyze_btn = gr.Button("Analyze", variant="primary")
-                        error_box = gr.Markdown("", visible=False)
-                    with gr.Column(scale=1):
-                        analysis_out = gr.Textbox(label="Generated Analysis", lines=18, interactive=False)
-                        gr.Markdown("**Chat with AURA**")
-                        chatbot = gr.Chatbot(height=420)
-                        user_input = gr.Textbox(placeholder="Ask follow-up...", label="Your question")
-                        send_btn = gr.Button("Send")
-                
-                analysis_state = gr.State("")
-                chat_state = gr.State([])
-
-                analyze_btn.click(
-                    fn=lambda txt: on_analyze(txt),
-                    inputs=[prompts],
-                    outputs=[analysis_out, error_box, analysis_state, chat_state]
-                )
-                send_btn.click(
-                    fn=lambda chat_list, msg, analysis_txt: on_send(chat_list, msg, analysis_txt),
-                    inputs=[chat_state, user_input, analysis_state],
-                    outputs=[chat_state, user_input]
-                )
-                user_input.submit(
-                    fn=lambda chat_list, msg, analysis_txt: on_send(chat_list, msg, analysis_txt),
-                    inputs=[chat_state, user_input, analysis_state],
-                    outputs=[chat_state, user_input]
-                )
-                chat_state.change(
-                    fn=convert_to_gradio_chat_format,
-                    inputs=[chat_state],
-                    outputs=[chatbot]
-                )
-
-            # MCP Server Tab
-            with gr.TabItem("MCP Server"):
-                gr.Markdown("**Call scraping and analysis functions directly:**")
-                with gr.Row():
-                    with gr.Column(scale=1):
-                        single_query = gr.Textbox(label="Single Scrape Query")
-                        scrape_btn = gr.Button("Scrape Query")
-                        scrape_out = gr.JSON()
-                        multi_queries = gr.Textbox(lines=6, label="Multi Scrape Queries")
-                        multi_scrape_btn = gr.Button("Multi Scrape")
-                        multi_scrape_out = gr.JSON()
-                    with gr.Column(scale=1):
-                        analysis_prompts = gr.Textbox(lines=6, label="Analysis Prompts")
-                        analyze_mcp_btn = gr.Button("Run Full Analysis")
-                        analyze_mcp_out = gr.JSON()
-                
-                scrape_btn.click(fn=mcp_scrape, inputs=[single_query], outputs=[scrape_out])
-                multi_scrape_btn.click(fn=mcp_multi_scrape, inputs=[multi_queries], outputs=[multi_scrape_out])
-                analyze_mcp_btn.click(fn=mcp_analyze, inputs=[analysis_prompts], outputs=[analyze_mcp_out])
+# ===================== GRADIO UI =====================
+def build_demo():
+    with gr.Blocks(title="AURA Chat — Hedge Fund Picks") as demo:
+        gr.HTML("""
+        <style>
+        .header {text-align:center; color:#2C3E50; font-size:32px; font-weight:bold; margin-bottom:10px;}
+        .container {background:#f9f9f9; border-radius:10px; padding:15px; margin-bottom:15px; box-shadow:0 4px 10px rgba(0,0,0,0.1);}
+        .instructions {color:#34495E; font-size:16px; line-height:1.6;}
+        .example {background:#EAF2F8; padding:8px; border-radius:5px; margin-top:5px; font-family:monospace;}
+        </style>
+        """)
+        gr.HTML('<div class="header">AURA Chat — Hedge Fund Picks</div>')
+        
+        # Explanatory YouTube video
+        gr.Video("https://youtu.be/56zpjyHd3d4", type="youtube", label="How it works")
+        
+        # Instructions container
+        gr.HTML("""
+        <div class="container instructions">
+        <b>What this does:</b> Fetches latest public data on insider trading and top stock market insights based on your prompts. 
+        It outputs top stock picks with <b>Investment Duration</b> guidance (when to buy and sell).<br><br>
+        <b>How to use:</b> Enter one or more prompts below, press <b>Analyze</b>, then chat with AURA about the results.<br><br>
+        <b>Example prompts you can copy:</b>
+        <div class="example">SEC insider transactions October 2025\n13F filings Q3 2025\nCompany: ACME corp insider buys</div>
+        <br>
+        The output will help you know which stocks are best to invest in and when to monitor alerts.
+        </div>
+        """)
+
+        # Main interface
+        with gr.Row():
+            with gr.Column(scale=1):
+                prompts = gr.Textbox(lines=6, label="Enter Prompts", placeholder="SEC insider transactions October 2025\n13F filings Q3 2025\nCompany: ACME corp insider buys")
+                analyze_btn = gr.Button("Analyze", variant="primary")
+                error_box = gr.Markdown("", visible=False)
+            
+            with gr.Column(scale=1):
+                analysis_out = gr.Textbox(label="Generated Analysis", lines=18, interactive=False)
+                gr.Markdown("**Chat with AURA about this analysis**")
+                chatbot = gr.Chatbot(height=420)
+                user_input = gr.Textbox(placeholder="Ask a follow-up question...", label="Your question")
+                send_btn = gr.Button("Send")
+
+        analysis_state = gr.State("")
+        chat_state = gr.State([])
+
+        def on_analyze(prompts_text):
+            analysis_text, initial_chat = analyze_and_seed_chat(prompts_text)
+            if analysis_text.startswith("ERROR"):
+                return "", f"**Error:** {analysis_text}", "", []
+            return analysis_text, "", analysis_text, initial_chat
+
+        def on_send(chat_state_list, user_msg, analysis_text):
+            if not user_msg.strip(): return chat_state_list or [], ""
+            updated_history = continue_chat(chat_state_list or [], user_msg, analysis_text)
+            return updated_history, ""
+
+        analyze_btn.click(fn=on_analyze, inputs=[prompts], outputs=[analysis_out, error_box, analysis_state, chat_state])
+        send_btn.click(fn=on_send, inputs=[chat_state, user_input, analysis_state], outputs=[chat_state, user_input])
+        user_input.submit(fn=on_send, inputs=[chat_state, user_input, analysis_state], outputs=[chat_state, user_input])
+        chat_state.change(fn=lambda msgs: msgs or [], inputs=[chat_state], outputs=[chatbot])
 
     return demo
 
-# =============================================================================
-# Handlers for Chat Tab
-# =============================================================================
-def on_analyze(prompts_text):
-    analysis_text, initial_chat = analyze_and_seed_chat(prompts_text)
-    if analysis_text.startswith("ERROR"):
-        return "", f"**Error:** {analysis_text}", "", []
-    return analysis_text, "", analysis_text, initial_chat
-
-def on_send(chat_state_list, user_msg, analysis_text):
-    if not user_msg.strip():
-        return chat_state_list or [], ""
-    updated_history = continue_chat(chat_state_list or [], user_msg, analysis_text)
-    return updated_history, ""
-
-# =============================================================================
-# CLEAN SHUTDOWN
-# =============================================================================
+# ===================== CLEAN SHUTDOWN =====================
 def _cleanup_on_exit():
     try:
         loop = asyncio.get_event_loop()
         if loop and not loop.is_closed():
             try: loop.stop()
-            except Exception: pass
+            except: pass
             try: loop.close()
-            except Exception: pass
-    except Exception: pass
+            except: pass
+    except: pass
 
 atexit.register(_cleanup_on_exit)
 
-# =============================================================================
-# RUN
-# =============================================================================
+# ===================== RUN =====================
 if __name__ == "__main__":
-    demo = build_demo_with_mcp()
+    demo = build_demo()
     demo.launch(server_name="0.0.0.0", server_port=int(os.environ.get("PORT", 7860)))