Vai al contenuto

Getting started

Quickstart

From zero to a populated read/write access graph in about five minutes, using the built-in demo estate. One static binary, no external services.

Questa documentazione è attualmente disponibile solo in inglese.

This is the fastest path to seeing what Olivares AI does: boot the single binary against its built-in demo estate and walk a real access graph end to end. No database to provision, no cloud account, no network egress.

Before you start

You need the controlplane binary. Grab a signed release artifact, or build it from source — it compiles to one static binary with the web console embedded, so there is nothing else to install. (The production install path is covered in the self-host guide; this page stays on the demo.)

Boot the demo estate

controlplane serve --seed-demo

On boot you’ll see a clearly marked DEMO MODE (synthetic data) banner and a printed URL with one-time demo credentials. Those credentials are public on purpose and bound to loopback — the demo is for local exploration, not a production posture.

Walk the graph

Open the console and follow the path the product is built around:

  1. Discover — the seeded agents and the resources they touch appear in the inventory.
  2. Read/write access map — every agent links to what it can reach, each edge typed read or read/write. This is the core view (see the access map concept, or the live product tour).
  3. Permitted vs Observed — the map diffs what each agent is allowed to touch against what it has been observed touching.

The demo estate is small and synthetic — a couple of dozen resource nodes and a similar number of access edges — and it deliberately seeds the two findings that matter: unexpected access (an agent reaching something it should not) and unused grants (permissions that are never exercised). The same serve --seed-demo boot is exercised by the product’s own smoke test, so these numbers don’t drift from the binary.

What’s real vs what’s demo

Everything you click is the real product — the same console, API and audit ledger you’d run in production. Only the estate data is synthetic. To build a graph of your own agents, connect a real source (for example PostgreSQL via pgAudit, or AWS via CloudTrail) instead of --seed-demo.

Next steps