Agent (holder) to controller Messaging Protocol

Background

There are a number of formal formalized protocols that are intended for agent-to-agent (e.g.: trust ping, coin flip, action menu). Thos protocols are processes at the agent level with some assistance from the controller.

There is a need for creating for business specific protocols, or even as a way to develop, experiment, and validate new protocols.

This protocol is intended to be implemented by the controller, and only a pass through on the agent. This is very similar to the basic message protocol with the main difference being that this exchange structured data, and it is not primarily intended for user interaction/consumption. It is primarily intended for the controller to process the messages.

Request (option 1)

{
    "@id": "123456780",
    "@type": "https://didcomm.org/rpc/1.0/request",
    "data": {
        {
            "@type": "https://didcomm.org/trust_ping/1.0/ping",
             "data":{
                "comment": "Hi. Are you listening?",
                "response_requested": true
             }
        }
     }
}

Request (option 2)

{
    "@id": "123456780",
    "@type": "https://didcomm.org/rpc/1.0/request",
    "@subtype": "https://didcomm.org/trust_ping/1.0/ping",
    "data": {
          "comment": "Hi. Are you listening?",
          "response_requested": true
     }
}

Response

{
    "@id": "123456780",
    "@type": "https://didcomm.org/rpc/1.0/response",
    "data": {
        {
            "@type": "https://didcomm.org/trust_ping/1.0/ping_response",
             "data":{
                "comment": "Hi yourself. I'm here."
             }
        }
     }
}

No implementation files, tests, or entry points are named. Start by locating the controller and agent pass-through/plugin entry points, then compare the proposed request and response envelopes with the linked basic message protocol. Done means the protocol shape and implementation scope are agreed and covered by appropriate tests.