refactor: mcp entrypoint changed

README.md

@@ -6,9 +6,147 @@ colorTo: purple
 sdk: gradio
 sdk_version: "5.49.1"
 app_file: app.py
-pinned: false
+pinned: true
 ---
 
 # Ragmint MCP Server
 
-Gradio-based MCP server for Ragmint.
+Gradio-based MCP server for Ragmint, enabling **Retrieval-Augmented Generation (RAG) pipeline optimization and tuning** via an MCP interface.
+
+<p align="center">
+  <img src="https://raw.githubusercontent.com/andyolivers/ragmint/main/src/ragmint/assets/img/ragmint-banner.png" width="auto" height="70px" alt="Ragmint Banner">
+</p>
+
+![Python](https://img.shields.io/badge/python-3.9%2B-blue) ![License](https://img.shields.io/badge/license-Apache%202.0-green) ![Status](https://img.shields.io/badge/Status-Active-success)
+
+---
+
+## 🧩 Overview
+
+Ragmint MCP Server exposes the full power of **Ragmint**, a modular Python library for **evaluating, optimizing, and tuning RAG pipelines**, through a **Multimodal Control Plane (MCP)**. This allows external clients (like Claude Desktop or Cursor) to **run experiments, retrieve leaderboard results, and tune RAG parameters programmatically**.
+
+### Features exposed via MCP:
+
+* ✅ Automated hyperparameter optimization (Grid, Random, Bayesian via Optuna)
+* 🤖 Auto-RAG Tuner for dynamic retriever–embedding recommendations
+* 🧮 Validation QA generation for corpora without labeled data
+* 🏆 Leaderboard tracking and experiment comparison
+* 🧠 Explainability via Gemini / Claude
+* 📦 Chunking, embeddings, retrievers, rerankers configuration
+* ⚙️ Full RAG pipeline control programmatically
+
+---
+
+## 🚀 Quick Start
+
+### Installation
+
+```bash
+pip install -r requirements.txt
+```
+
+### Running the MCP Server
+
+```bash
+python app.py
+```
+
+The server will expose MCP-compatible endpoints, allowing clients to:
+
+* Perform optimization experiments
+* Automatically autotune pipelines.
+* Generate validation QA sets with LLM.
+
+
+### Environment Variables
+
+Set API keys for LLMs used in explainability and QA generation:
+
+```bash
+export ANTHROPIC_API_KEY="your_claude_key"
+export GOOGLE_API_KEY="your_gemini_key"
+```
+
+---
+
+## 🧠 MCP Usage
+
+Ragmint MCP Server provides Python-callable interfaces for programmatic control. Example usage with MCP:
+
+```python
+from mcp_client import MCPClient
+
+client = MCPClient(server_url="http://localhost:7860")
+
+# Run Auto-RAG tuning
+config, results = client.autotune(docs_path="data/docs/", trials=5)
+print("Best config:", config)
+
+# Retrieve leaderboard
+top_results = client.leaderboard(top_k=5)
+print(top_results)
+```
+
+---
+
+## 🔤 Supported Embeddings
+
+* `sentence-transformers/all-MiniLM-L6-v2`
+* `sentence-transformers/all-mpnet-base-v2`
+* `BAAI/bge-base-en-v1.5`
+* `intfloat/multilingual-e5-base`
+
+### Configuration Example
+
+```yaml
+embedding_model: sentence-transformers/all-MiniLM-L6-v2
+```
+
+---
+
+## 🔍 Supported Retrievers
+
+| Retriever    | Description                        |
+| ------------ | ---------------------------------- |
+| FAISS        | Fast vector similarity search      |
+| Chroma       | Persistent vector DB               |
+| scikit-learn | Local lightweight NearestNeighbors |
+
+### Configuration Example
+
+```yaml
+retriever: faiss
+```
+
+---
+
+## 🧮 Dataset Options
+
+| Mode                 | Example                            | Description                                  |
+| -------------------- | ---------------------------------- | -------------------------------------------- |
+| Default              | validation_set=None                | Uses built-in experiments/validation_qa.json |
+| Custom File          | validation_set="data/my_eval.json" | Your QA dataset                              |
+| Hugging Face Dataset | validation_set="squad"             | Downloads benchmark dataset                  |
+
+
+---
+
+## 🧩 Folder Structure
+
+```
+ragmint_mcp_server/
+├── app.py  # MCP server entrypoint
+├── models.py
+└── api.py
+```
+
+---
+
+## 📘 License
+
+This project is licensed under the Apache License 2.0. See the [LICENSE](LICENSE) file for details.
+
+
+<p align="center">
+  <sub>Built with ❤️ by <a href="https://andyolivers.com">André Oliveira</a> | Apache 2.0 License</sub>
+</p>

api.py

@@ -303,7 +303,6 @@ def generate_qa(req: QARequest):
         raise HTTPException(status_code=500, detail=str(exc))
 
 
-# only run uvicorn if script is executed directly
-if __name__ == "__main__":
-    import uvicorn as _uvicorn
-    _uvicorn.run(app, host="0.0.0.0", port=7860, log_level="info")
+def start_api():
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=8000, log_level="info")

app.py

@@ -1,15 +1,17 @@
-# app.py
 import gradio as gr
 import requests
 import json
 import os
 import shutil
-import uvicorn
 from models import OptimizeRequest, AutotuneRequest, QARequest
-from api import app as backend_app  # import the FastAPI app we just saved
+import threading
+from api import start_api
 
-# Base URL for internal calls (same process)
-BASE_INTERNAL = "http://127.0.0.1:7860"
+threading.Thread(target=start_api, daemon=True).start()
+
+
+# Base URL for internal calls
+BASE_INTERNAL = "http://127.0.0.1:8000"
 
 def call_api(endpoint: str, payload: dict) -> str:
     try:
@@ -46,13 +48,13 @@ DEFAULT_AUTOTUNE_JSON = model_to_json(AutotuneRequest)
 DEFAULT_QA_JSON = model_to_json(QARequest)
 
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# Ragmint MCP Client (UI)")
+    gr.Markdown("# Ragmint MCP Client")
     with gr.Column():
         gr.Markdown("## Upload Documents")
         upload_files = gr.File(file_count="multiple", type="filepath")
         upload_path = gr.Textbox(value=DEFAULT_UPLOAD_PATH, label="Docs Path")
         upload_btn = gr.Button("Upload", variant="primary")
-        upload_out = gr.Textbox()
+        upload_out = gr.Textbox(label="Response")
         upload_btn.click(upload_docs_tool, inputs=[upload_files, upload_path], outputs=upload_out)
         gr.Markdown("---")
 
@@ -60,7 +62,7 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
         gr.Markdown("## Optimize RAG")
         optimize_input = gr.Textbox(lines=12, value=DEFAULT_OPTIMIZE_JSON, label="OptimizeRequest JSON")
         optimize_btn = gr.Button("Submit", variant="primary")
-        optimize_out = gr.Textbox(lines=15)
+        optimize_out = gr.Textbox(lines=15,label="Response")
         optimize_btn.click(optimize_rag_tool, inputs=optimize_input, outputs=optimize_out)
         gr.Markdown("---")
 
@@ -76,13 +78,13 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
         gr.Markdown("## Generate QA")
         qa_input = gr.Textbox(lines=12, value=DEFAULT_QA_JSON, label="QARequest JSON")
         qa_btn = gr.Button("Submit", variant="primary")
-        qa_out = gr.Textbox(lines=15)
+        qa_out = gr.Textbox(lines=15,label="Response")
         qa_btn.click(generate_qa_tool, inputs=qa_input, outputs=qa_out)
         gr.Markdown("---")
 
-# mount the Gradio app on FastAPI at root ("/")
-gr.mount_gradio_app(backend_app, demo, path="/")
-
-# When run directly, serve with uvicorn (HF will run this)
 if __name__ == "__main__":
-    uvicorn.run(backend_app, host="0.0.0.0", port=7860, log_level="info")
+    demo.launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+        mcp_server=True
+    )