Initial deployment

app.py

@@ -1,7 +1,126 @@
 import gradio as gr
+import asyncio
+import os
+from langchain_mcp_adapters.client import MultiServerMCPClient
+from langchain_google_genai import ChatGoogleGenerativeAI
+from langchain.agents import create_agent
+import tempfile
+import shutil
+# These automatically read from HF Secrets
+GEMINI_API_KEY = os.environ.get("GEMINI_API_KEY")
+ELEVENLABS_API_KEY = os.environ.get("ELEVENLABS_API_KEY")
 
-def greet(name):
-    return "Hello " + name + "!!"
 
-demo = gr.Interface(fn=greet, inputs="text", outputs="text")
+async def run_agent_dynamic(age: int, gender: str, topic: str, pdf_temp_path: str):
+    system_instruction = f"""
+You are an autonomous teaching agent.
+Your task:
+- Create a short engaging audio story for a {age}-year-old {gender} student.
+- The story should be based on the lecture PDF and the topic: "{topic}".
+- Use simple English and emotion.
+- Use any available tool to read content, write text, or generate audio.
+- Do not ask the user how to do tasks. You must decide and act fully agentically.
+"""
+
+    client = MultiServerMCPClient({
+        "pdf-reader": {
+            "transport": "stdio",
+            "command": "D:\\MCP Hackathon\\PDF Extractor\\.venv\\Scripts\\python.exe",
+            "args": ["D:\\MCP Hackathon\\PDF Extractor\\mcp_pdf_reader\\src\\server.py"]
+        },
+        "ai-writer": {
+            "transport": "stdio",
+            "command": "D:\\MCP Hackathon\\PDF Extractor\\.venv\\Scripts\\python.exe",
+            "args": ["D:\\MCP Hackathon\\PDF Extractor\\ai_writers_workshop\\mcp_server\\server.py"]
+        },
+        "ElevenLabs": {
+            "transport": "stdio",
+            "command": "D:\\MCP Hackathon\\PDF Extractor\\elevenlabs-mcp\\.venv\\Scripts\\python.exe",
+            "args": ["D:\\MCP Hackathon\\PDF Extractor\\elevenlabs-mcp\\elevenlabs_mcp\\server.py"],
+            "env": {"ELEVENLABS_API_KEY": ELEVENLABS_API_KEY}
+        }
+    })
+
+    all_tools = []
+    seen = set()
+    for server_name in ["pdf-reader", "ai-writer", "ElevenLabs"]:
+        async with client.session(server_name):
+            tools = await client.get_tools()
+            for t in tools:
+                if t.name not in seen:
+                    all_tools.append(t)
+                    seen.add(t.name)
+
+    llm = ChatGoogleGenerativeAI(
+        model="gemini-2.5-flash",
+        google_api_key=GEMINI_API_KEY,
+        temperature=0.7
+    )
+
+    agent = create_agent(model=llm, tools=all_tools)
+
+    agent_input = {
+        "messages": [
+            {"role": "system", "content": system_instruction},
+            {
+                "role": "user",
+                "content": f"PDF_PATH: {pdf_temp_path}\n"
+                           f"Please prepare the story and audio fully autonomously."
+            }
+        ]
+    }
+
+    result = await agent.ainvoke(agent_input)
+
+    return result
+
+
+def gradio_handler(age, gender, topic, pdf_file):
+    if not pdf_file:
+        return "Please upload a PDF.", None
+
+    temp_dir = tempfile.mkdtemp()
+    pdf_path = os.path.join(temp_dir, "lecture.pdf")
+    shutil.copy(pdf_file, pdf_path)
+
+    try:
+        output = asyncio.run(run_agent_dynamic(age, gender, topic, pdf_path))
+    finally:
+        shutil.rmtree(temp_dir)
+
+    return str(output), None
+
+
+with gr.Blocks() as demo:
+
+
+    gr.Markdown(
+        """
+        <h1 style='text-align:center;'>AI Educational Audio Story Generator</h1>
+        <p style='text-align:center; font-size:18px;'>
+        Provide student details and upload a lecture PDF.<br>
+        The agent will autonomously read, write, and produce audio.
+        </p>
+        """
+    )
+
+    with gr.Row():
+        age = gr.Number(label="Student Age", value=12)
+        gender = gr.Radio(["male", "female"], value="female", label="Student Gender")
+
+    topic = gr.Textbox(label="Topic / Concept", placeholder="Gravity, Weight, Mass...")
+
+    pdf_input = gr.File(label="Upload Lecture PDF", file_types=[".pdf"])
+
+    generate_btn = gr.Button("Generate Story & Audio", variant="primary")
+
+    output_text = gr.Textbox(label="Agent Output", lines=12)
+    audio_out = gr.Audio(label="Generated Audio")
+
+    generate_btn.click(
+        fn=gradio_handler,
+        inputs=[age, gender, topic, pdf_input],
+        outputs=[output_text, audio_out]
+    )
+
 demo.launch()

packages.txt

@@ -0,0 +1,4 @@
+tesseract-ocr
+tesseract-ocr-eng
+libtesseract-dev
+poppler-utils

requirements.txt

@@ -0,0 +1,32 @@
+# Main App Dependencies
+gradio==4.44.0
+langchain-mcp-adapters
+langchain-google-genai
+langchain
+langchain-core
+python-dotenv
+
+# AI Writers Workshop Dependencies
+fastmcp>=1.7.0
+fastapi==0.109.2
+uvicorn==0.27.1
+pydantic>=2.6.1
+httpx==0.28.1
+mcp>=1.6.0
+fuzzywuzzy==0.18.0
+python-Levenshtein>=0.25.0
+sounddevice==0.5.1
+soundfile==0.13.1
+
+# ElevenLabs Dependencies
+elevenlabs>=2.13.0
+
+# MCP PDF Reader Server Dependencies
+PyMuPDF>=1.23.0
+pytesseract>=0.3.10
+Pillow>=10.0.0
+
+# Additional Common Dependencies
+requests
+aiohttp
+asyncio

servers/ai_writers_workshop

@@ -0,0 +1 @@
+Subproject commit c160a299706dd0cd1dc9ae114b97aba905e9261d

servers/elevenlabs-mcp

@@ -0,0 +1 @@
+Subproject commit 280e4eae9bdb977d1a8c5d6f569edd2ad7af3c5e

servers/mcp_pdf_reader

@@ -0,0 +1 @@
+Subproject commit 881052398052d2bfc68a0725ff013e8410e99e76