HF Space: add metadata + root requirements

README.md

@@ -1,3 +1,12 @@
+---
+title: Agent MCP SQL
+emoji: 🧠
+sdk: streamlit
+app_file: space_app.py
+python_version: 3.11
+pinned: false
+---
+
 # GraphRAG Agentic System
 
 ## Overview
@@ -88,6 +97,34 @@ The agent will then:
 4. Execute the query using the `QueryExecutorTool`.
 5. Return the final answer to the UI.
 
+### Deploying a Hugging Face Space (Streamlit front-end only)
+
+This repo includes a self-contained Streamlit app for Hugging Face Spaces: `space_app.py`.
+It connects to your externally reachable Agent and MCP services.
+
+1) Expose your services (public host or tunnel)
+   - Agent FastAPI endpoint: `https://<your-host>/query`
+   - MCP FastAPI base: `https://<your-host>/mcp`
+
+2) In a new HF Space (Streamlit), add these files:
+   - `space_app.py` (entrypoint)
+   - `requirements.txt` with:
+     ```
+     streamlit==1.28.0
+     requests==2.31.0
+     pandas==2.1.0
+     ```
+
+3) In Space Settings → Variables and secrets, set:
+   - `AGENT_URL` (e.g., `https://your-agent-host/query`)
+   - `MCP_URL` (e.g., `https://your-mcp-host/mcp`)
+   - `MCP_API_KEY` (the MCP auth key)
+   - (Optional) `AGENT_HEALTH_URL`, `NEO4J_URL`
+
+4) Configure the Space to run `space_app.py` as the Streamlit app file.
+
+Once the Space starts, it will display the same chat UI and stream responses from your hosted Agent.
+
 ## Development
 
 ### Running the Agent Manually
@@ -124,4 +161,8 @@ To test the agent's logic directly without the full Docker stack, you can run it
 ├── docker-compose.yml
 ├── Makefile
 └── README.md
-```
+```
+
+
+testing, what do you see from the mcp and db?
+

requirements.txt

@@ -0,0 +1,3 @@
+streamlit==1.28.0
+requests==2.31.0
+pandas==2.1.0