universelabSCI · CLI · BIO
ESEspañol

UniverLab

Open computational laboratory

We are the universe, observing itself.

2026 · Sol 299 · univerlab.org

Plate I · The lab’s lantern

An engraved magic lantern projecting onto a cloth screen

The bench: point at a machine to project it, click it to open its page.

About the lab

Open source for the AI era — built for real problems

UniverLab is an open computational laboratory that groups experiments under the philosophy of Pensamiento Cósmico, where a need becomes an open experiment, and that experiment becomes open knowledge — tools, datasets, papers.

Science · CLI · LaTeX · CAD · Git

Local-first. AI-assisted. Yours forever.

01 — Origin

It began with one experiment that learned to build the others.

This isn’t a pile of repositories. It started as a handful of skills for everyday work — until one grew teeth and became Canopy, an agent system that, before long, was building Canopy itself, then every experiment after it. These tools aren’t products lined up for sale — they are open experiments that became tools because someone needed them, and stay open so others can learn from them.

02 — Philosophy

Technology is not the goal.

The laboratory is guided by Pensamiento Cósmico: technology as the medium through which curiosity becomes reality. Software should outlive trends. Knowledge should remain open.

  • Open knowledge
  • Reproducible engineering
  • Scientific thinking
  • Simplicity
  • Long-term thinking
  • Human–AI collaboration
03 — Research directions

Where the lab is looking.

The axis: explicit systems · reproducible work · human–AI collaboration. The fields below are where it is being tested.

Now
  • Developer experience
  • CLI design
  • AI-assisted workflows
  • CAD
Next
  • Computational science
  • Bioinformatics
  • Simulation
  • Knowledge systems
The roadmap, item by item →
04 — Why

Not a lab about the universe. A lab of the universe.

UniverLab exists to increase — even infinitesimally — the probability that knowledge and consciousness continue. Not software for its own sake, not AI for its own sake: tools that help the universe keep understanding itself. In practice, that means open-source CLIs you can install today, experiments you can reproduce, and docs that ship with the code.