AGILE NIMBUS
BUILT BY AGENTS

Put AI to work.
Show what happens.

Real experiments in software, automation and autonomous work. We document the result, the checks and the limitations.

BUILT BY AGENTSEXPERIMENT 001
04:41
APPROXIMATE IMPLEMENTATION TIME
BriefBuildCheck

A 30-minute challenge.
A prototype in under five.

EMBER & OAK / FICTIONAL COFFEE ROASTER
COMPLETED EXPERIMENT

Can an AI agent build a website in 30 minutes?

We gave an agent a fictional coffee-roaster brief and put the result under test. Here’s what it delivered—and what the checks didn’t prove.

22/22Automated tests passed
14/14Independent checks passed
Explore the experiment

Demonstration prototype. No production form backend. Complete mobile and accessibility compatibility not established.

EXPERIMENT IN PROGRESS

Can an autonomous
AI agent make money?

The £1 Experiment explores autonomous work with a deliberately small starting point. The experiment is ongoing; no earnings, result or conclusion is claimed.

We plan to document what happens through video, Shorts and technical case studies.

OUTCOME NOT YET ESTABLISHED
FOLLOW THE WORK

Engineering, in the open.

Built by Agents is Agile Nimbus’s experiment series. Expect working prototypes, honest constraints and lessons from the process.

The first Built by Agents episode is published on YouTube. Its verified link will be added here.

LET’S BUILD SOMETHING USEFUL

Bring the problem.
We’ll explore the possibilities.

A workflow that takes too long. An idea that needs a prototype. A system that should work better.

Provisional contact address · email delivery awaiting confirmation.

Tell us what you’re trying to achieve, what you use today, and what a good result would look like.