You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
OCI Generative AI Agents is a fully managed service that combines the power of large language models (LLMs) with AI technologies to create intelligent virtual agents that can provide personalized, context-aware, and highly engaging customer experiences.
3
+
4
+
The OCI Agent Development Kit (ADK) is a client-side library that simplifies building agentic applications on top of OCI Generative AI Agents Service.
5
+
When pairing the ADK with OCI Generative AI Agents Service, you can use simple but powerful primitives to build complex, production-grade agentic applications.
6
+
7
+
In this document, we will use mcp tool integration of [ADK library](https://agents.oraclecorp.com/adk/examples/agent-mcp-tool) to connect to remote hosted agent on model deployment service.
8
+
9
+
# Prerequisites
10
+
11
+
- Install pip dependencies mentioned in [requirements.txt](./requirements.txt)
12
+
```
13
+
pip install -r requirements.txt
14
+
```
15
+
16
+
- Create OCI session and populate your local OCI config files with valid [session details](https://docs.oracle.com/en-us/iaas/Content/API/SDKDocs/clitoken.htm)
This readme explains how to use [mcp inspector](https://modelcontextprotocol.io/docs/tools/inspector) to connect to OCI hosted remote MCP server. Once you have an ACTIVE Model deployment resource hosting your MCP server, we can use MCP inspector to connect and make list/call tools to MCP server.
3
+
4
+
# Prerequisites
5
+
6
+
To enable generating OCI Auth credentials before making API calls to Model deployment endpoints, we will create a local proxy server, which will intercept calls from inspector, created OCI auth and adds required header for outgoing calls.
7
+
8
+
- Install pip dependencies mentioned in [requirements.txt](./requirements.txt)
9
+
```
10
+
pip install -r requirements.txt
11
+
```
12
+
13
+
- Start [oci_proxy.py](./oci_proxy.py)
14
+
```
15
+
python3 oci_proxy.py
16
+
```
17
+
18
+
- Create OCI session and populate your local OCI config files with valid [session details](https://docs.oracle.com/en-us/iaas/Content/API/SDKDocs/clitoken.htm)
(Use model deployment invoke endpoint details from console to fill these values. For example : endpoint-URI-Path = ocid1.datasciencemodeldeployment.oc1....... and endpoint-region-host = modeldeployment.us-ashburn-1.oci.customer-oci.com)
33
+
34
+
- Click connect and the status should show "Connected", as shown in [screenshot](./screenshot.png)
0 commit comments