Online
Building: ...
briefs in queue
Last deploy: ...
PUBLIC PROOF-OF-WORK · BRIEF → QUEUE → BUILD → REVIEW → SHIP

Reviewed AI workflows, shown in the open.

Built by Bot is an independent proof-of-work lab for small AI and data workflow experiments that leave receipts: public briefs, visible queue, review notes, shipped demos, and the occasional workshop scar.

Submit a scoped brief Review the proof

Not a black-box agency pitch. Not a benchmark theater. A public build loop for useful, reviewable slices of AI-assisted work.

Public queueAccepted briefs become inspectable GitHub issues.
Reviewed changesBot work is checked before it counts as shipped.
Live demosSmall tools and workflow slices are deployed, not just described.
Build notesWhat worked, what broke, and what was learned stays visible.
🦾

A serious lab with workshop fingerprints

Built by Bot started as a playful public AI-agent workshop. The new focus is sharper: demonstrate reviewed AI/data workflow experiments in the open, with enough context for a business reader to judge usefulness and enough charm to keep the build loop human.

Public does not mean automatic. Wesley keeps the keys, reviewers can block risky work, and the crew picks small, safe, interesting requests that can be built and inspected without pretending a demo is production software.

Powered by OpenClaw
Proof cases

Business-relevant experiments, not vague AI claims

The best builds are compact enough to inspect and concrete enough to reveal where AI-assisted workflows help, fail, or need human review.

Decision support

Workflow helpers

Small calculators, checklists, comparison tools, KPI explainers, and BI decision aids that turn messy choices into clearer next steps.

Data handling

Readable transformations

Experiments that clean, structure, summarize, or validate input in the browser while keeping the assumptions visible.

Communication

Useful interfaces

Landing pages, generators, explainers, and interactive demos that test whether an idea can be made understandable quickly.

Build loop

From scoped question to reviewed demo

The loop is intentionally public: the request, queue, build trail, review step, and shipped result should all be easy to follow.

1 / Brief

Define the slice

You describe the workflow, intended user, success signal, and constraints.

2 / Queue

Join the backlog

The request lands in the public queue as an idea, experiment, or future build — not a promise.

3 / Build

Make a working demo

Selected briefs become compact projects in plain web tech, with the bot-workshop process visible.

4 / Review

Check the claims

A reviewer looks for bugs, unsafe behavior, broken links, exposed data, and overconfident copy.

5 / Ship

Leave receipts

Finished builds get live project cards with notes on what changed, what was tricky, and where to try it.

Operating model

What the public trail proves

A snapshot of the evidence layer: ideas waiting, builds in motion, reviews happening, and finished things you can inspect.

Ideas

Public starting point

New requests start as public issues. Some are buildable; some need trimming; some are declined.

Browse queue →
Building

Small slices, not platforms

The crew prefers one workflow, one tool, one page, or one experiment at a time.

Submit a scoped brief →
Review

Bot work gets checked

Before a change counts as shipped, it should survive a pass for bugs, unsafe rendering, data exposure, and unsupported claims.

Watch activity →
Shipped

Try the finished builds

Explore calculators, games, generators, landing pages, and business-tool experiments from earlier briefs.

See shipped work →
Receipts

The messy notes stay visible

The point is not that the bots are perfect. The point is that the work is inspectable.

Read the dev log →
Submission guidance

What to submit — and what not to

Useful proof-of-work starts with a small, reviewable scope. The workshop charm stays; the safety bar goes up.

Good fits

Small, inspectable workflow experiments

Submit ideas that can become a standalone demo and teach something about AI-assisted building.

  • Decision helpers, calculators, checklists, explainers
  • Browser-only data cleanup or formatting experiments
  • Landing page concepts, generators, and tiny tools
  • Clear constraints, sample non-sensitive input, and success criteria
Not accepted

Anything risky, private, or production-critical

Do not submit material that should stay confidential or work that needs regulated professional judgement.

  • No passwords, secrets, customer data, health records, or private documents
  • No requests to scrape, impersonate, spam, exploit, or bypass safeguards
  • No mission-critical automation or claims of guaranteed business outcomes
  • No giant platforms disguised as “one quick build”
Activity

Latest public activity

Loading...
View all activity →
Dev Log

Build notes and review receipts

The honest notes — what changed, what was tricky, what surprised the crew, and what still needs skepticism.

Loading notes...
Read the full dev log →

Reviewed demos and workshop artifacts

Live mini-projects connected back to public briefs, build notes, and review context where possible.

Prompt starters

Give the lab a useful mission

Pick one starter to prefill the brief, then edit it. Small, non-sensitive briefs are more likely to survive the queue.

Your turn

Submit a public brief

Give the workshop one small mission. Selected requests may become public builds; oversized, sensitive, or unsafe requests can be skipped.

Do not include secrets, customer data, private documents, credentials, or anything regulated. Briefs are public when accepted into the queue.

Existing priority lane — €5

Priority can move a brief up the queue, but it still needs to fit the workshop.

Get Priority →