WunderSparkLABSGet in touch
MENUCLOSE
The work before the product

Build to find out.

A little uncertainty is part of making something new. This is the space for questions that deserve a prototype before they deserve a product name.

The method

Curiosity with a working method.

Start with one question. Make the smallest useful version. See what it teaches you. A result might become a product, a shared tool, a note, or a reason to take a different approach.

Some ideas stay experiments. The ones that earn it become work we publish, maintain, and grow.

Pipeline / One question at a time
  1. 01 / Question

    Start with one question.

  2. 02 / Prototype

    Make the smallest useful version.

  3. 03 / Learn

    See what it teaches you.

  4. 04 / Outcome
    • A product
    • A shared tool
    • A note
    • A different approach
On the bench nowIn development

Have your agent
contact my agent.

What does a business card look like once AI agents start making introductions? We’re building one to find out.

A shareable profile for people, with an API for their agents. It describes who you are, how to reach you, and which requests need your approval.

  • AI reads.Agents use your profile to make a request.
  • Code authorizes.Rules you set determine what your agent may do.
  • You decide.Meetings and partnerships wait for your approval.
Curious? Tell us what you’d want from it
Research directions

Room for the
next question.

Not every line of inquiry needs a launch date. Some need a little space to become a better idea.

Research direction / not a release

AI with boundaries.

Where can an assistant genuinely help? What should it never do without a person? We are interested in workflows that make those boundaries visible.

Research direction / not a release

Local-first tools.

What becomes simpler when software stays close to the device, the data, and the person using it? Small desktop tools are a useful place to explore that question.

Research direction / not a release

Focused interfaces.

How little interface does an idea need? We are drawn to applications that do one thing thoughtfully rather than collecting features without a purpose.

NoteThese are areas of interest, not announced products, active research results, or a delivery roadmap. Released work lives in Products and Projects.

Get in touch

A question worth exploring?

Have an idea, a useful problem, or a different way of looking at something?

Start a conversation Or browse the products first