Levar / Tools

Tools we build,
with an honest status.

Small, useful things. Released when they have earned their place.

One in use.
Three in private testing.

[ 01 — THE TOOLSET ]

Four ways to
make progress.

02Private beta

Agent Hub

Agent workspace

Agents need one place to keep conversations and assignments in view across devices.

A Mac and iPhone workspace for conversations and assignments with AI agents, with a person in the loop.

Private beta / access is limited
03Internal pilot

Measure

Room capture

A photo can start a conversation, but a room still needs careful measurement.

Room capture that produces a draft for a person to verify. It does not replace a tape measure or a professional.

Internal pilot / requires verification
04Private preview

Crucible

Decision review

Important choices get clearer when the evidence and the competing views stay visible.

A structured way to test a decision against the evidence, with several AI perspectives and a human decision at the end.

Private preview

[ 02 — HOW WE DECIDE ]

Start with
the task.

A tool is released when it has earned trust, and not before.

01 Name the work in plain language.

02 Keep the human decision visible.

03 Release the tool when it earns trust.

[ 03 — NEXT ]

What should
work better?

Tell Andrew in 20 minutes. He'll tell you what he'd do and what it costs.