What Toneinux does, in the order it does it

Every board starts with a sheet your team already keeps. From the workspace you add a connector, grant it read access to that sheet, and choose the tabs it may look at. The connector holds a read scope only: it can open the sheet, and it cannot change a cell, add a tab or leave a comment. Your sheet's own permissions stay the last word, so revoking the connector's access at the sheet's end cuts the board off at once.

The connector reads on a schedule you set, every few minutes for an operations board or once a day for a monthly one, and it reads on demand when anyone presses refresh. Each read is recorded with its time, which is what lets every panel say when its figure was last true.

The board does not read cell addresses; it reads named ranges. In the sheet you select the block that matters, a single total, a column of weekly figures, a small table, and give it a name. In the workspace that name appears in the list of ranges the connector can see, and you tick the ones the board should read.

Naming does two things. It tells the board which cells matter, so a panel ignores the rest of the tab. And it binds the panel to meaning rather than position: insert a row above the block and the range moves with it, so the panel keeps reading the right cells. Names are also the vocabulary the team shares. A range called revenue_week is something everybody can talk about; a cell address is not.

A board is a page with a small number of panels on it. Each panel is drawn from one named range and is one of four kinds. A figure shows a single value, large, with the range name and its last read beneath it. A trend draws a line through a row or column of values, oldest to newest. A table shows a rectangular range as it is, with the sheet's own headers. A note shows a text cell, so the person who owns a number can write the sentence that goes with it.

You place panels on a grid, resize them and reorder them by dragging. The board keeps one arrangement per board, not one per person, so everybody reads the same layout, and a viewer opening the link sees exactly what the editor placed.

Every board has one link inside your workspace. Anyone holding a seat can open it; nobody without a seat can. Seats come in two kinds. An editor can place, move and remove panels and bind ranges. A viewer opens the board, reads it and can step back through snapshots, but cannot move a thing. Most teams give editor seats to the two or three people who own the sheet and viewer seats to everyone who reads it.

There are no public links and no embed codes, so a board cannot leak by being pasted somewhere. A link handed to someone without a seat opens a sign-in page and nothing else.

Read it as it stood

Each refresh stores the values it read as a snapshot. Pick a date on the board and every panel shows the value from the snapshot nearest that moment, with the read time printed under it. Snapshots are kept for the life of the workspace and leave with you as plain files when you close it.

The board does no arithmetic of its own. If you need a figure that does not exist in the sheet, add it to the sheet. That is where formulas belong: visible, owned by the people who maintain them, and picked up by the board on the next read without anyone having to configure a thing.