Skip to main content
Glama
README.md
# Project Title

Turn your REST APIs to an MCP server

## Description

This sample runs a IBM Granite model using ollama locally. The code uses gradio to expose an MCP server with a tool to call the POST API and get the response in JSON

## Getting Started

### Dependencies

* Need python 3.12 or greater 

* Need ollama and here are the instructions to run IBM Granite on ollama running on mac os

* [instructions] (https://www.ibm.com/granite/docs/run/granite-with-ollama/mac/)

### Test

```
curl -XPOST -H "Content-Type: application/json" http://localhost:11434/v1/chat/completions -d '{"model": "granite3.2:8b", "messages": [{"role": "users", "content": "How are you today?"}]}'

```

### Executing program

```
python -m venv apiasmcp
source apiasmcp/bin/activate
pip install -r requirements.txt
export GRADIO_MCP_SERVER=True
python app.py
```

### Access locally

Open http://localhost:7860 in your browser
Enter some text and click “Submit”
You should see the results


![Test](runlocally.png)

### Add it to an MCP client

Visit http://localhost:7860/gradio_api/mcp/schema
This shows the MCP tool schema that clients will use
You can also find this in the “View API” link in the footer of your Gradio app
Add this JSON BLOB to your MCP client:

```
{
  "mcpServers": {
    "gradio": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://localhost:7860/gradio_api/mcp/sse"
      ]
    }
  }
}

```

For example in cursor mcp client, you should see as below

![Cursor](cursormcpgradio.png)