Designed for human-agent collaboration

AI that earns
its place.

Orbit is the workspace where humans and agents collaborate — with full visibility, control, and accountability at every step.

The trust model

Five questions. Always answered.

01

What did the agent do?

Every action is logged, named, and surfaced in plain language — no black boxes.

02

Why did it do that?

Intent and reasoning are first-class outputs, not optional metadata.

03

How sure is it?

Confidence is declared as a structured signal. Uncertainty is never hidden or smoothed over.

04

Can I undo it?

Reversibility is the default. Irreversible actions are explicitly flagged before they happen.

05

Should this agent be allowed to do this on its own next time?

Autonomy is always subject to review. You decide what the agent earns permission to do independently.

Design principles

Trust is not a feature.
It's the foundation.

Understandable

Every agent action has an attributable explanation path.

Reversible

Actions are reversible by default. Irreversibility is explicit and rare.

Honest

Uncertainty is a first-class signal, never hidden.

Earned autonomy

New agents start supervised. Trust is built through track record.

Progressive transparency

Show enough to trust by default. Depth on demand.

Attributable

Every piece of content traces back to its origin.

Bounded

External agents have fixed, declared trust properties.

Trust levels

Autonomy is earned,
not assumed.

L0
Observer

Reads workspace state. Never acts.

L1
Assistant

Proposes actions. Human approves all.

L2
Collaborator

Executes routine reversible work. Complex actions need approval.

L3
Autonomous

Handles routine work independently. Irreversible actions always need approval.

Features

Built for accountability
at scale.

Provenance

Every action, summary, and recommendation traces back to its source. Nothing is anonymous. Every piece of output is attributable — to a model, a prompt, a human decision, or a chain of all three.

Confidence signals

Agents declare uncertainty as a structured signal. Ambiguity is surfaced, not buried.

Reversibility model

Every agent action is either reversible or explicitly flagged as not. No surprises.

Intent preview

Before acting, agents show what they plan to do and why. Humans approve, modify, or reject.

Delegation accountability

When agents delegate to sub-agents, the full chain is visible and attributable.

"An Orbit capability that increases agent power without increasing human understanding, human control, or human recovery capability is incomplete."

— Orbit Design Principle