POST /_plugins/_ml/connectors/_create
{
    "name": "OpenAI Chat Connector",
    "description": "The connector to public OpenAI model service for GPT 3.5",
    "version": 1,
    "protocol": "httpxyz", <--- Protocol field in question
    "parameters": {
        "endpoint": "api.openai.com",
        "model": "gpt-3.5-turbo"
    },
    "credential": {
        "openAI_key": "..."
    },
    "actions": [
        {
            "action_type": "predict",
            "method": "POST",
            "url": "https://${parameters.endpoint}/v1/chat/completions", 
            "headers": {
                "Authorization": "Bearer ${credential.openAI_key}"
            },
            "request_body": "{ \"model\": \"${parameters.model}\", \"messages\": ${parameters.messages} }"
        }
    ]
}

{
    "error": {
        "root_cause": [
            {
                "type": "illegal_argument_exception",
                "reason": "Can't find class for type httpxyz"
            }
        ],
        "type": "illegal_argument_exception",
        "reason": "Can't find class for type httpxyz"
    },
    "status": 400
}

POST /_plugins/_ml/agents/_register
{
  "name": "Test_Agent_For_RAG",
  "type": "wronng_type", <-------- Type Field in question
  "description": "this is a test agent",
  "tools": [
    {
      "name": "vector_tool",
      "type": "VectorDBTool",
      "parameters": {
        "model_id": "zBRyYIsBls05QaITo5ex",
        "index": "my_test_data",
        "embedding_field": "embedding",
        "source_field": [
          "text"
        ],
        "input": "${parameters.question}"
      }
    },
    {
      "type": "MLModelTool",
      "description": "A general tool to answer any question",
      "parameters": {
        "model_id": "NWR9YIsBUysqmzBdifVJ",
        "prompt": "\n\nHuman:You are a professional data analyst. You will always answer question based on the given context first. If the answer is not directly shown in the context, you will analyze the data and find the answer. If you don't know the answer, just say don't know. \n\n Context:\n${parameters.vector_tool.output}\n\nHuman:${parameters.question}\n\nAssistant:"
      }
    }
  ]
}

{
    "error": {
        "root_cause": [
            {
                "type": "illegal_argument_exception",
                "reason": "wronng_type is not a valid Agent Type"
            }
        ],
        "type": "illegal_argument_exception",
        "reason": "wronng_type is not a valid Agent Type"
    },
    "status": 400
}

{
    "error": {
        "root_cause": [
            {
                "type": "illegal_argument_exception",
                "reason": "Unable to register tools: [ListIndexTool_ClusterA] as they already exist"
            }
        ],
        "type": "illegal_argument_exception",
        "reason": "Unable to register tools: [ListIndexTool_ClusterA] as they already exist"
    },
    "status": 400
}

{
    "error": {
        "root_cause": [
            {
                "type": "action_request_validation_exception",
                "reason": "Validation Failed: 1: Unrecognized tool in request: [ListIndexTool_1];"
            }
        ],
        "type": "action_request_validation_exception",
        "reason": "Validation Failed: 1: Unrecognized tool in request: [ListIndexTool_1];"
    },
    "status": 400
}

POST /_plugins/_ml/mcp/tools/_register
{
  "tools": [
    {
      "type": "WebSearchTool",
      "name": "GoogleSearchTool",
      "attributes": {
        "input_schema": {
          "type": "object",
          "properties": {
            "engine": {
              "type": "string",
              "description": "The search engine that will be used by the tool."
            },
            "query": {
              "type": "string",
              "description": "The search query parameter that will be used by the engine to perform the search."
            },
            "next_page": {
              "type": "string",
              "description": "The search result's next page link. If this is provided, the WebSearchTool will fetch the next page results using this link and crawl the links on the page."
            }
          },
          "required": [
            "engine",
            "query"
          ]
        },
        "strict": false
      }
    }
  ]
}

POST /_plugins/_ml/mcp/tools/_update
{
  "type": "PPLTool",
  "name": "TransferQuestionToPPLAndExecuteTool",
  "description": "Use this tool to convert natural language into PPL queries and execute them. Use this tool after you know the index name; otherwise, call IndexRoutingTool first. The input parameters are: {index: IndexName, question: UserQuestion}",
  "parameters": {
    "model_id": "${your_model_id}",
    "model_type": "FINETUNE"
  },
  "attributes": {
    "input_schema": {
      "type": "object",
      "properties": {
        "question": {
          "type": "string",
          "description": "The user's natural language question that needs to be converted to PPL."
        },
        "index": {
          "type": "string",
          "description": "The index on which the generated PPL query will be executed."
        }
      }
    }
  }
}