AI Interface

The selfdriven AI Interface provides a unified interface for accessing, orchestrating, and extending AI-driven capabilities across the selfdriven.network ecosystem.

It serves as the foundational layer that connects human identity, community context, and intelligence through a consistent, secure, and verifiable surface.

The Interface exposes a suite of methods for managing models, agents, conversations, vector stores, and assistants.

Each method is accessed via authenticated HTTP POST requests, with all request and response data transmitted as JSON over SSL for privacy and integrity.

By aligning with the Progressive Self-Actuation Framework, the Interface enables communities, organisations, and practitioners to embed AI within trusted, context-aware environments — where every interaction is authenticated, auditable, and human-aligned.

The Interface supports all widely recognised foundation models and integrates with OpenRouter to provide access to specialised models.

Version:

2.3 (August 2025)

Endpoints:

ai.interface.selfdriven.network
ai.api.slfdrvn.io

Core Methods

Category	Method	Description
Models	ai-gen-get-models	Retrieve available AI model specifications.
Agents	ai-gen-get-agents	List or describe active conversational agents.
Conversation	ai-gen-conversation-chat, ai-gen-chat	Initiate or manage AI conversation threads.
Vector Stores	ai-gen-util-vector-store-create, ai-gen-util-vector-store-attach-file, ai-gen-util-vector-stores	Create and manage contextual knowledge bases.
Files	ai-gen-util-file-upload, ai-gen-util-file-base64-upload, ai-gen-util-files	Securely upload and manage files for embedding or reference.
Assistants	ai-gen-util-assistant-create, ai-gen-util-assistants	Define and manage assistant configurations.
Threads	ai-gen-util-thread-create, ai-gen-util-thread-chat	Manage conversational threads for context continuity.
Utilities	ai-gen-util-service-models	Access underlying AI service metadata.

Protocol Overview

All methods use HTTP POST.
Requests and responses use application/json format.
All communications are protected via SSL/TLS.
Standard HTTP status: 200 OK
Internal status field: “status”: “OK” or “ER”
All responses contain a “method” and “data” object for structured integration.

Example

Endpoint: https://ai.interface.selfdriven.network
Method: ai-gen-get-models

Request:


{
  "apikey": "e7849d3a-d8a3-49c7-8b27-70b85047e0f1",
  "authkey": "28cc4fae-804f-4603-d08a-94fce2be90f2",
  "mode": { "type": "live" },
  "data": { "version": "1" }
}

Methods

ai-gen-get-models
ai-gen-get-agents
ai-gen-conversation-chat
ai-gen-chat
ai-gen-util-service-models
ai-gen-util-vector-store-create
ai-gen-util-vector-stores
ai-gen-util-vector-store-attach-file
ai-gen-util-file-upload
ai-gen-util-file-base64-upload
ai-gen-util-files
ai-gen-util-assistant-create
ai-gen-util-assistants
ai-gen-util-thread-create
ai-gen-util-thread-chat

Notes

All methods are access as http POSTs
All data to be sent in the body as JSON - so protected by SSL.
All response data returned as JSON.
All http response statuses are 200 with status=”OK,ER”.

Methods

Get Models (ai-gen-get-models)

Mode	Current set to “reflect” - ie request is validated and all data sent with request reflected back for integration development/testing.
Reflect Options	status: eg “OK”, “ER” data: Data to be reflected.
Request Data	method: ‘get-specs’ apikey: [supplied] authkey: [supplied] data: version:
Format	JSON
Response Data	method: ‘ai-gen-get-models’ status: ‘OK’, ‘ER’ - if error then {error: {code:, description:}} data: [ { name:, url:, id:, etag: modifieddatetime:, createddatetime: } ]

Conversation Chat (ai-gen-conversation-chat)

Mode	Current set to “reflect” - ie request is validated and all data sent with request reflected back for integration development/testing.
Reflect Options	status: eg “OK”, “ER” data: Data to be reflected.
Request Data	method: ‘ai-gen-conversation-chat’ apikey: [supplied] authkey: [supplied] data: messagesystem: message:
Format	JSON
Response Data	method: ‘ai-gen-conversation-chat’ status: ‘OK’, ‘ER’ - if error then {error: {code:, description:}} data: [ { firstname:, lastname:, email: , id:, etag: modifieddatetime:, createddatetime: } ]