Your backend as a connector. Add it once, use it everywhere Claude runs.
Every Norbix installation is an MCP server — the hosted address below, nothing to install. The tool signs in with OAuth in your browser and gets described tools marked read or change.
norbix login → norbix ai init → norbix … --json --dry-run first, --yes only for destructive changes. Same commands in CI.https://hub.norbix.ai/v3/account/mcp The blocks below are the ones Cloud shows for the managed service. On your own installation Cloud fills in your own addresses.
Claude on the web and in the desktop app share the same connectors: add it once, use it in both. Cowork — Claude's desktop agent — sees it too.
https://hub.norbix.ai/v3/account/mcpIn Claude open Customize → Connectors, click + Add, then Add custom connector. Name it Norbix and paste the MCP address below as the remote MCP server URL — with https, with /v3, without a slash at the end.
Keep Use Claude's published identity (Register automatically works too). Leave Request headers empty. Click Add, then Connect.
Your browser opens Cloud's consent screen: pick the roles Claude gets and press Allow. Back in a chat, click + → Connectors and switch Norbix on.
Claude Desktop's config file starts local servers, so it reaches the hosted server through the mcp-remote bridge.
{
"mcpServers": {
"norbix": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://hub.norbix.ai/v3/account/mcp"]
}
}
}Settings → Developer → Edit Config. Restart Claude Desktop; the consent screen names the tool MCP CLI Proxy (the bridge).
Only an owner can add a custom connector: Organization settings → Connectors → Add → Custom → Web, same name, URL and OAuth client. Members then find it under Customize → Connectors and click Connect. The Free plan allows one custom connector.
Claude calls a read tool, for example get_database_taxonomies, and turns the JSON into a list. Ask for a change and the tool either goes through (Administrator) or answers "Caller is missing required permission" (Viewer).
Show the taxonomies of my Norbix project.
Create a project, enable modules, build collections from a schema bundle and publish them.
Find, count, insert and update records; taxonomies, aggregates and triggers on every collection.
Email, push, SMS and payment providers; templates, campaigns, files and storage.
Roles and policies, sign-in integrations, invitations and the membership settings.
Triggers across modules, scheduler tasks, serverless code functions and webhooks.
Logs by correlation id, diagnostics, usage — the same read-only look Cloud gives you.
OAuth opens Cloud's consent screen. You tick the projects and roles the tool gets — Viewer, Developer, Data Editor, Administrator or your own.
The tool acts as an AI service user. Remove it under Account → AI service users and it is signed out.
Every tool is marked read or change. Tools that send messages, delete or change a lot answer requires_confirmation first.
Every call the tool makes is in Logs with its correlation id, like any other request.
Open your avatar → Connect your AI in Cloud. It holds a prompt you can paste into any AI tool, the Hub, API and MCP addresses of your installation, and one tab per tool with the exact commands — including the key variants for CI.
The address, OAuth or a key, the consent screen, roles and environments.
Copy-paste setup for every tool, a first prompt and what it returns.
Every tool group per module, which ones read, change, or ask first.
The Viewer and Auditor roles, and a key for CI.
Create a free account, connect Claude, and ask it for your first backend change.