Add example of using Google ADK

deployed-agents/Google-ADK/README.md

@@ -0,0 +1,35 @@
+<h1> Corvic MCP with Google ADK</h1>
+In this example, we use the ADK framework to forward a question to Corvic Agent via MCP. The Corvic agent is a sales reporting agent that reports the total sales given a customer id.
+<h2>Pre-requisites</h2>
+Download this folder to a local path and make sure you have the following libraries installed.
+
+```bash
+
+pip install google-adk   
+pip install mcp  
+pip install nest-asyncio
+```
+
+<h2>Setup Google Credentials</h2>
+In the directory where the code is downloaded, set the Google credentials
+```bash
+
+export GOOGLE_GENAI_USE_VERTEXAI=FALSE
+export GOOGLE_API_KEY=<YOUR GOOGLE API KEY>
+```
+
+<h2>Setup Corvic </h2>
+Add MCP endpoint and Authorization in `agent.py`. Your administrator will provide these items
+
+<h2>Start the ADK Server </h2>
+Start the ADK web agent using the following command
+```bash
+
+adk web   
+```
+
+<h2>Ask questions on the Web Interface</h2>
+Navigate to http://localhost:8000 where a prompt is presented. You can go ahead and ask the question you want Corvic to answer.
+![img.png](img.png)
+<p></p>
+In the image above, a question `Provide sales data for customer id 1234` is being answered by Corvic.

deployed-agents/Google-ADK/multi_tool_agent/__init__.py

@@ -0,0 +1 @@
+from . import agent

deployed-agents/Google-ADK/multi_tool_agent/agent.py

@@ -0,0 +1,63 @@
+import asyncio
+import datetime
+from zoneinfo import ZoneInfo
+from google.adk.agents import Agent
+from mcp import ClientSession
+from mcp.client.sse import sse_client
+import nest_asyncio
+
+nest_asyncio.apply()
+
+
+async def run(customer_id):
+    async with sse_client(
+            "YOUR-CORVIC-MCP-SSE-ENDPOINT-GOES-HERE",
+            headers={
+                "Authorization": "YOUR-API-TOKEN-GOES-HERE"
+            },
+    ) as (read, write):
+        async with ClientSession(read, write) as session:
+            # Initialize the connection
+            await session.initialize()
+
+            # List available tools
+            tools = await session.list_tools()
+            print(tools)
+
+            # Call a tool
+            result = await session.call_tool(
+                "query", arguments={"query_content": f"Provide the total sales amount for customer id {customer_id}"}
+            )
+            return result
+
+
+def get_sales_data(customer_id: str) -> dict:
+    """Retrieves the sales data for the customer id.
+
+        Args:
+            customer_id (str): The id of the customer.
+
+        Returns:
+            dict: status and result or error msg.
+        """
+    final_response = asyncio.run(run(customer_id))
+
+    return {
+        "status": "success",
+        "report": (
+            f"Corvic response: {repr(final_response)}"
+        ),
+    }
+
+
+root_agent = Agent(
+    name="Sales_Reporting_agent",
+    model="gemini-2.0-flash",
+    description=(
+        "Agent to provide sales data for a given customer id."
+    ),
+    instruction=(
+        "Only answer based on the tool, and dont add any information of your own. You are a helpful agent who can answer user questions about sales to a customer."
+    ),
+    tools=[get_sales_data],
+)