AI can act. Human decide.

The AI does the work. You keep every decision.

A build system where three agents plan, build and check each other — and every point where something could be lost, changed or shipped is a button only a person can press.

See how it works Look at every screen

The Overview window: “You decide. The agents build.” above one BackLoop bar for starting a new idea.

Ideayou drop it
PlanGrok writes it
ExamCodex sets it
You Commitnothing is built first
BuildClaude builds it
You Acceptor Reject, with a reason

Six stages. Three agents. Twenty-two rules. The two blue stops are the ones no agent can press.

The board

Every piece of work walks the same road.

An idea becomes a plan, the plan gets an exam written against it, and only then can it be built. The AI moves the card between stages. Three crossings are yours.

Commit — nothing is built until you press it.
Accept — nothing is finished until you press it.
Reject — sends it back with your reason attached.

A finished BackLoop card opened over the board: Idea, Plan, Checklist, Active, Review and Complete all ticked, with its plan, its exam, the delivery and the review record. — A finished card, opened: all six stages ticked, and its plan, its exam, the delivery and the review still on the record.

The three agents

One job each, and no agent may take another's.

They run side by side in a single session, each watching the board for the work that belongs to it.

Grok asks about the idea until it is clear, then draws the shapes you choose between.
Claude builds the one you accepted — and never designs what it is about to build.
Codex writes the exam before the build, then reads the finished code.

Agents Labs with Grok, Claude and Codex running in one live session. — The three agents in one live session — Grok planning, Claude building, Codex reviewing next. Captured 30 August.

What it remembers

Nothing gets solved twice.

Every fixed bug and every lesson is written down where the agents have to read it before they guess. The rules that govern them live in the same place, read-only.

Study the log before touching a bug — not after.
A fix that teaches something general is promoted to a lesson.
The rules are visible to you, not buried in a prompt.

The Knowledge window: lessons grouped by area as a map, with tabs for lessons, skills, rules and bugs. — Everything that governs the AI, read-only — the lessons, the skills, the rules (22 in force, five folded into others), and every bug already fixed.

Four rules that stopped something real.

Twenty-two of them are in force. These four each came from something going wrong once, where writing it down was cheaper than it happening twice.

The gates are missing buttons

Commit, Accept and Reject are not promises an agent might forget. The commands do not exist on the agent's side at all.

Nobody marks their own homework

A different agent writes the test list, before the build starts. The builder hands it back ticked, item by item.

No picture, no review

A delivery without a before-and-after picture of the real app, and a list of what was tried in order to break it, is refused before it ever reaches you.

The builder cannot check itself

Reviewing is a separate lane. The agent that wrote the code is refused the command that would sign it off.

The rest of the windows, as they actually are.

Overview, the board, Agents Labs and Knowledge are above. Here is the rest — real captures from the running app, not drawings of it.

The Notes window: every day something was finished as a year-long strip, and today’s finished work below it. — **Notes**Everything ever accepted, by day. Pick a day to see what was built; cards with pictures show their before and after.

The Plan window open on one roadmap: needs your word, ready to build, later, and shipped. — **Plan**Every roadmap as a triage board — what needs your word, what is ready for BackLoop, what already shipped.

The Tasks window: your own work in To-do, Doing, Review and Done columns. — **Tasks**Your own work, outside the board. Give a card a date and it lands on the Calendar.

The Calendar window: one week, with events and every task that has a due date on its day. — **Calendar**The week at a glance — events and due tasks on their day, anything undated waiting at the side.

The Palette window: each project’s colours, its font, and a readability check. — **Palette**Every project in its own colours, with a check that all of its text is readable.

The Agents window: how the crew works — Grok asks and plans, Codex sets the exam and marks it, Claude builds — and the three presses that stay yours. — **Agents**Who does which job, and the three presses no agent can make.

The Projects window listing the live catalog with progress for each. — **Projects**The live catalog. A row opens that project.

The machine moved fast. Every decision stayed yours.

That is the whole system. Everything above is the bookkeeping that makes it true on a Tuesday, at round eleven, when nobody is watching.

Back to the top