Documentation | AI Nexus

Start here

Three steps to a working workspace, whether you stay in the WebApp or call the API.

Open the WebApp, pick your login mode, and land on Dashboard or Assistant.

Admins connect at least one model provider and a working default model.

Chat in Playground, or send POST /api/ai/query with an API key.

How-tos for the product UI. See the full list under Guides in the navbar.

The owner-facing assistant, missions, and reviewable global memory.

Build, deploy, and supervise agents from templates or from scratch.

Upload documents, retrieve passages, and ground answers in your corpus.

Wake agents on Slack, Shopify, or other events — not only on a schedule.

Connect Claude, Cursor, or VS Code to your AI Nexus memory over MCP.

Members, keys, and the project boundary most product work lives in.

HTTP contract for integrators: auth, AI endpoints, MCP, agents, and webhooks.

Base URLs, first request, and how the HTTP surface is organized.

JWT, API keys, and what each caller is allowed to do.

Query, chat, and /v1 routes that speak the OpenAI protocol.

Hosted tools, connections, and the personal memory server.