The Messaging API is designed to help you perform operations related to messaging. These include managing agents, creating chats, sending messages, and receiving streamed responses in real time.

The API is organized around REST (opens in a new tab) for resource operations, and adds a realtime (socket.io) connection so that responses from agents can be streamed back to your app as they are generated. Our goal is to make the API predictable, with resource-oriented URLs and standard HTTP response codes and verbs.

It exposes only the messaging surface of ProteusAI — agents, chats, messages, and repositories. For the full platform (orgs, plugins, workflows, app integrations, etc.) use the Management API.

Base URL

https://messaging-api.useproteus.ai

All REST endpoints are mounted under /api (e.g. https://messaging-api.useproteus.ai/api/chats).

How it works

Sending a message dispatches it to the agents in the chat. An agent generates a reply and posts it back; the Messaging API relays that reply — optionally streamed chunk-by-chunk — to every client connected to the chat's realtime room.

• Use REST to manage agents, create chats, read message history, and manage repositories.
• Use the realtime connection (or the SDK) to send messages and receive replies, including streamed responses.

All responses use the envelope { "ok": boolean, "data"?: ..., "error"?: { "message": string }, "meta"?: ... }.

Client Libraries

npm install --save @proteus-ai/sdk

See the Node SDK for the easiest way to send messages and consume streamed responses.