Docs

How Langrove works

Every app is built from the same blocks. Your AI agent puts them together, and you can see every part in the console.

  • Projects

    One project is one app: its data, code, sign-in, rules and website.

  • Data

    Tables your agent creates and changes. Test data stays apart from live data.

  • Code

    Node.js that runs in your account. Each endpoint says who may call it.

  • Access rules

    Who may read or change each record. Everything is denied by default.

  • Sign-in

    Email codes or links, passwords, passkeys, or Microsoft, Google and any OIDC.

  • Files and knowledge

    Files kept with their records, and documents people and agents can search.

  • AI agents and workflows

    Started by a message, a schedule or an event, and never reaching further than the person behind it.

  • Secrets

    Set by a person and never readable, by your agent or anyone else.

  • Website

    Your app's own pages, at the same address as its backend.

  • Where it runs

    Langrove's AWS in the EU, or your own AWS account. Your app keeps answering with Langrove switched off.

Around the console

Overview
Status, release, usage and a map of the system
Health
Requests, errors and response times
Logs
Every request, for live or test data
Access
Sign-in, keys, secrets and connections
Trust
What holds, and what AI editors can and can't do
Settings
Connect your AI, domains, rename or delete

Next: Releases and environments