In progress

AGINUS

Arvor's multi-purpose agent, being built in the open

AGINUS is the vision of a single agent that understands context, remembers what matters and actually acts — evolving with Memento as its memory backbone. It isn't ready yet. We're building it in public, with no date promised.

No date promised. Real progress, documented in public.

Follow the build in public

The vision

One agent, many purposes.

Today every task needs a different tool. AGINUS wants to be the agent you call for anything — and one that gets better with every conversation, because it carries real memory instead of starting from zero.

01

Perception

Understanding multi-channel context: text, voice, image — whatever the task calls for.

02

Memory

Memento as the backbone. What AGINUS learns today, it carries into the next conversation.

03

Action

Not just a chat that replies. An agent that executes real tasks, with supervision.

04

Transparency

Built in public. Progress, technical decisions and limitations, documented as they happen — no date promised.

Want to know when AGINUS opens up?

No spam, no date promised. Just a note when there's something real to show.

Join the waitlist