Documents

A Document is a full rich text document you create in MotherFluxer — generate one from a chat prompt or upload a file you already have, then write, format, proof, and export it. Documents live under the Productivity filter group in Artifacts.

Create a Document

You have two ways to start a Document:

• Generate from chat. Describe what you want in the chat — a report, a letter, an outline, a study guide — and MotherFluxer writes a complete Document you can keep editing.
• Upload a file. Bring in a document you already have. Supported uploads include PDF, DOCX, PPTX, TXT, and similar text formats. MotherFluxer imports the content into an editable Document. A Document has one page size for all its pages, so a PDF whose pages are not all the same size comes in at its first page's size.

Either way, the result lands in your Artifacts under the Productivity filter so you can find it again, reopen it, and keep working. Its preview picture is drawn the first time you open it, so a Document you have not opened yet shows a file icon in Artifacts until then.

Open the document editor

Open any Document to launch the document editor. Across the top you'll find three mode tabs:

• Smart Edit — describe a change in plain language and let MotherFluxer rewrite the whole Document for you.
• The tool menus on the left — Style, View, Add and Layers. Take the controls yourself to format, write, annotate and proof by hand.
• File — save and revisit versions, share, and export.

Move between them freely as you work — write a draft by hand, hand it to Smart Edit for a pass, then fine-tune it yourself.

Your edits are saved as you go, a moment after you pause, so you don't have to close the editor to keep them. A saved version is a different thing: one you deliberately mark from the File menu so you can come back to it later.

Smart Edit a Document

On a Document, Smart Edit is more than one tool — it's the Document's full set of AI tools, and you reach all of them by describing what you want in Discuss:

• A whole-Document rewrite — "make this more formal," "tighten it to one page," "rewrite for a younger audience" — regenerates the entire Document in one pass, so it's the fastest way to reshape tone, structure, or length at once. Anything the rewrite doesn't touch — your page setup, header and footer — is kept as it was.
• One-tap edits on a selected element — Improve, Shorten, Elaborate and Change tone, applied to just the heading, paragraph or text box you picked rather than the whole page.
• AI reviews that leave your text alone — inline critique, an overall assessment with scores, and a side-by-side comparison of several models' feedback.
• Research & write — give the AI a subject and it researches the web, then writes a structured Document with citations back to its sources.
• Layout tools — tidy the alignment and spacing, reflow the content into a named template, turn it into a presentation, or reorder the sections.
• AI placeholders — mark a spot while drafting and have the AI fill every marked spot in later.
• Change settings — tune elements already on the page from a sentence — a look, where one stands, its edge — or the paper's size, margins and colour. It never rewrites your words.

Each of these saves its result as a new version, so your previous one isn't lost, and each shows its cost before it runs.

A Document edit can be shown pictures as well — a layout to match, or a chart to write about. Add them in the row under the edit's model choice: the paperclip brings one in from your device, and Pick image chooses one you already have.

An edit set to One page or One element instead of Whole document lists the pages and elements of the Document as it stands on your screen, including anything you added since it was last saved.

The edits Smart Edit can run on a Document, each with what it does:

• Edit document — Change the document from your own instruction — the whole thing, one page, or one element
• Improve — Rewrite for clarity and flow, and fix spelling, grammar and punctuation
• Shorten — Make it more concise without losing meaning
• Elaborate — Expand with more detail and supporting points
• Change tone — Rewrite in a tone you describe
• Auto-layout — Realign elements, fix spacing, and remove overlaps
• AI critique — Add inline review comments you can act on or dismiss
• Assess — Get an overall review with scores and suggestions
• Research & write — Research a topic on the web and write a fully-cited document
• Multi-model review — Review with 3 AI models side-by-side and compare
• Apply template — Reflow the document into a named layout (report, two-column, résumé, …)
• Make presentation — Convert the document into a slide-style deck (one section per slide)
• Rearrange — Reorder sections and elements into a more logical flow
• Change settings — Change an element's or the paper's settings from a sentence

Edit a Document by hand

The editor gives you hands-on control through four menus:

• Style — the core rich text editor for typing, styling, headings, lists, and inline formatting, and the page's colour. The text look card at its right puts a look on the words you select. Capitals (the Aa beside B · I · U · S) sets the selected words in capitals, and Outline, beside the words' colour, draws a line round them — so a look's capitals or outline can be taken off again here too. The same card stands at the right of a table's Table tab, for its words.
• View — set the page up and proof it: page size, margins, orientation, the wide or page view, guides, spellcheck, grammar, the dictionary and find & replace. Spellcheck and Grammar each read how many issues they found; press one and a small navigator steps you through them, with the check's On/Off switch at its top. Pinch the page (or hold Ctrl and scroll) to zoom it, and the readout in the top bar puts it back to fit.
• Add — put something new on the page: text, a shape, an emoji, a table, a freeform stroke, a picture, a 3D model, a mask or a blur. Picking one drops it on the page and closes the menu; its own settings then appear along the bottom of the screen. Freeform drops an empty drawing with its Freeform menu open, and it works as in the Image editor: you are drawing straight away, and its brushes, its connector and each brush's Recipe — Stamps, Taper and Lands, where any brush can be set to Erase — are the same, as is what a pen's stroke keeps of how hard it pressed, how far it leaned and when it passed. Image and 3D Model add your newest picture or 3D model at once, and the item's first tab along the bottom is where you swap it for another of yours; if you have none yet, Browse opens on its upload instead. A picture comes in at its full quality, the same as in the Image and Video editors. Browse, the last tile, holds more places to look: an upload from your device, your own artifacts, and Starred and community, the items and groups you have starred and the community has posted — the same Browse the Image and Video editors have. What other people have published is in the search on the item's first tab, under Community.
• Layers — reorder what sits in front, and put several elements into a group that moves, turns, flips and resizes as one. Press and hold on the page and Layers opens narrowed to what lies under your finger, with a Filtered card at its start to clear it.

Move between these tabs as needed — draft in Style, set the page up and clean it up in View, add to it from Add, and stack it in Layers.

The editor draws your items on your device's graphics processor, so it opens in the same browsers as the Image editor: anywhere else, opening a Document shows This artifact can't be opened on this device. with the same reason beneath it as an Image does. The Document's card still shows. Its public page shows too if it holds only words; with shapes, pictures with effects, drawings or other items on it, the public page says the same thing, because those are drawn by the graphics processor as well.

Settings live along the bottom, with the same tabs an Image has. Click anything you have added and its settings open there. The first tab is named for what the item is — Text, Shape, Table, Image, Mask and so on — and holds how it looks (for a picture, which of your pictures it shows). Effects holds the same looks as in an Image and the same Form section — Repeat for copies of the item, Deform to bend it with a brush — though not an Image's Blend card: an item on a page always lands over what is beneath it. Border holds its edge (and, for a picture, the shape it is cut to), and Transform holds its crop (for a picture, a shape, an emoji, a drawing or a 3D model), flip and opacity — with Perspective and, on a picture, Splice, as in the Image editor — how the page's text flows around it and how far the text keeps from it, and last, in one row, where it sits, how big it is and how it is turned. A 3D model's first tab is 3D, with Effects, Border and Transform beside it (see below), a blur and a privacy cover have their first tab and Transform alone, and a mask can follow a person or thing in a picture on its page through Tracking on its Mask tab. A text box or table wearing an effect shows with the effect on it, and goes back to its plain words while you type in it. Word and PowerPoint receive such a box as a picture. Clicking the open tab again puts the item down. You can click straight from one thing to another at any time — it does not matter which menu is open at the top.

Picking on the page. A tap or click selects the item showing at that spot, and the same press can drag it straight away: once the pointer travels a few pixels the item moves, while a slightly shaky tap only selects it. Tap the selected item again to select the one underneath it at that spot, as on every editor. An item in a group selects its whole group first, and a second tap steps into the group, to the item under your finger; once you're inside, a tap on another item of the same group selects that item directly. A selected text box or table is the exception: a second click types into it, with the cursor where you clicked — or, on a table, in the cell you clicked — wherever nothing else is drawn over that spot. A click anywhere on a text's speech bubble — inside it, on its outline or on its tail — selects the text, and so does a click on a line's full thickness or on a connector's arrowheads. An outline-only shape is selected on its outline, not in its empty middle, and a drawing on its strokes, not in the gaps between them; a few pixels around every item still count. The box around a selected item takes in its bubble and tail. The arrow keys nudge the selected item a pixel at a time (ten with Shift), and with Alt held the left and right arrows turn it.

A selected item stays in its place in the stack. If something else covers part of it, the selected item's box and handles are drawn on top, but the covering item still shows over it, just as it did before you selected it. Dragging where the covering item is moves that item, not the selected one.

The buttons above the box. A selected item has a blue outline with white corner dots and side handles to resize it — a corner dot keeps the item's proportions, a side handle stretches it one way; text, a group and a 3D object take only the corner dots, a table its corners and the handles on its left and right sides — and one row of buttons above it: the rotate button (drag it to turn the item; while you hold it the angle shows at the bottom of the screen), Duplicate, Delete, and the ⋯ More menu: Lock, then Fit display (the item as big as fits whole on the page, centred, keeping its turn and tilt — a text box's words grow or shrink with it; one Undo puts it back), then a row of three text-wrap pictures — none, top and bottom, and wrap — that set how the page's text flows around the item (the current one is lit), then a row of four layer moves — bring to front, bring forward, send backward and send to back. Nothing sits under the box. When the item fills the screen, the row sits just inside its top edge.

Locking. Lock, in a selected item's ⋯ More menu, locks it. A locked item can't be moved, resized, turned or deleted: it shows just its outline, and the row above it is Duplicate and More, with Unlock in More. Try to move it and a lock appears over it for a moment, then fades. In Layers, a locked item's card shows a small lock at its left edge. A group with a locked item in it is locked as a whole, and Lock in the group's More menu locks or unlocks all of its items at once.

Typing, drawing and the controls on an item. Click a selected text box again to type into it, or click a cell of a selected table — or press Enter. Press a brush in a drawing's Freeform menu, or Draw at its right end, to draw into it, and click a connector on the page to work on its points. A shape's reshape grips — an arrow's head and shaft, a triangle's peak, a star's points and the rest, the same as on an Image — and a speech bubble's tail sit right on the selected item as small blue rings, and a 3D model's turn as a larger ring at its centre, whenever the item is big enough on screen — zoom in on a small one to bring them back. While you type, move points or hold a ring, the box's handles and the row above it step aside; Escape, a click elsewhere or leaving the menu that opened it ends typing and point editing. Draw is the exception: while you draw, the box and its buttons are hidden and the Freeform menu's right end reads Drawing, with an Exit.

Typing a number. The number boxes on the Transform tab (where the item sits, its Size, its Rotation, and how far wrapped text keeps from it) take effect when you press Enter or click away. They don't change on every key, so typing 150 never passes through 1 and 15 first. The up and down arrow keys step a number instead; hold Shift to step by ten. Each number you type is one Undo step, and a run of arrow presses on the same box is one step too. The font size box in a header's or footer's window works the same way. Open where the item sits and a small square beside its numbers picks which corner they read — press it to step round the four corners; a Size you type keeps that corner where it is on the page, even on a turned item. A width you type for a text box scales its words with it, as on an Image or a Video.

Headers and footers. Click a page's header or footer to type into it, and a small window opens beside it with four rows: the font and its size; B, I and Capitals; the words' colour and an Outline, with its width once it is on; and Insert field, which puts a page number, the page count, a date or a time where your cursor is. A text look card stands at the rows' right: press it to see every look, and the one you pick goes on the header's or footer's type. Escape or a click elsewhere closes the window.

Change the page size, orientation or margins

The View menu sets the page up: its size, portrait or landscape, and its margins. Custom on either one opens a small window where you type your own numbers, applied when you press Done. The text that flows down the page rewraps to the new width. Every item placed on the page — a text box, a picture, a shape, an emoji, a drawing, a table, a blur — keeps its shape: it follows the page to the same part of it and grows or shrinks to fit, with everything in it (its text, its outlines and line widths) growing or shrinking with it, so nothing is ever stretched or squashed. A group moves and resizes as one piece, so its items keep their spacing. Change the setting back and every item returns to exactly where it was, at exactly its old size.

While the page has different settings from the ones it started with, every number you see is the size you see: a stroke width, a font size or a position on the Transform tab reads what the page shows, and a value you type is what the page will show.

The page's own settings

View

• Page size — The paper's width and height in the page's pixels, 96 to an inch: every item keeps its place on it.
• Margins — The paper's margins in the page's pixels — the top, the bottom, and the two sides alike: where the flowing text stands.

Style

• Page color — The colour of the paper, under the margins and the flowing text.

Stack items and groups

In the document editor, what sits in front is set by the stacking order. You change it with the layer row at the bottom of the ⋯ More menu above a selected item (bring to front, bring forward, send backward, send to back), and the cards in Layers. A group always moves through that order as one piece:

• Move an item forward or back past a group and it passes the whole group in one step. It never lands between the group's members.
• A group moved forward or back passes the next item or group as a whole, in either direction.
• An item that belongs to a group moves only within its group.

A selected group's first tab is its Group menu — the same one an Image and a Video give a group: one card per member, named as the Layers strip names it, with the member's leave-the-group and copy buttons when you select it, and at the right end a container with Ungroup all and Consolidate as its rows (Consolidate draws the members into one picture that takes their place; one Undo brings the group back) and, under a thin line, a delete button and the forward and backward arrows, which act on the member you picked or, with none picked, on the whole group — a star leading them then, which stars the group whole. Its Effects, Border and Transform tabs are beside it, the same ones a group has in an Image.

A move that would change nothing is dimmed, on a Layers card and in the More menu's layer row alike, and each move is one Undo step.

Select a group and its Transform settings show the group's own box: where it sits, its Size, its Rotation and its opacity. You can type into them like any item's, and a typed size keeps the group's proportions. Turn a group and its box turns with it, staying snug around its members instead of growing into a level box around them. The turn is one Undo step.

Export a Document

Export from the File menu. A Document can be saved out in several formats:

• DOCX
• PPTX
• PDF
• PNG
• HTML
• Markdown
• Text

Pick the format you need and MotherFluxer prepares the file for download.

The list ends with Add to an App…: pick one of your Apps there, and the format you choose next goes into that App's Assets instead of downloading.

Your file is built right on your device and downloads immediately — there's no server wait and no link that expires. If the file can't be built, Export failed appears in the middle of the screen, and your Document is left as it was. If part of the Document — an emoji, a shape, a blur, a drawing or another picture — can't be drawn on your device, you'll see Part of this document could not be drawn, so no file was made. with Try again. No file is ever made with a piece missing.

A Document the AI made as a PDF, Word or PowerPoint file downloads the same way: Download builds the file on your device from the Document, pictures included.

A label for AI-made content. When AI wrote the Document, or added to it, the PDF, DOCX, PPTX and PNG files carry a hidden, machine-readable label saying it contains AI-made material, kept in the file's own information — nothing is added to the pages you see. A Document made only from your own writing and edits carries no label, and once AI has added to a Document its files keep the label even if you later take that part out. Text, Markdown and HTML have no place to keep such a label, so they carry none.

3D models and pictures. A 3D model exports as the picture of it its box shows, in every format. If you export while a 3D model's redraw is still running, the export waits for it to finish, so the file always has the latest picture. Every picture comes across in PDF, DOCX and PPTX whatever format it was saved in. A cropped item shows only the part you kept in every format; a cropped picture in PPTX stays a whole picture cropped by PowerPoint, so you can re-crop it there, and in PDF, DOCX and PPTX every item of a cropped group comes across as a picture of what it shows.

PDF, PNG, HTML and PPTX keep every page as you see it. The text you type on the page flows across every page as it does in the editor, around items you've set text to wrap around, with your header and footer on each page. Every item sits where you put it with the turn, mirroring and see-through setting you gave it, and items that overlap stack in the same order as in the editor. The files carry the Document's fonts, so they look the same on any computer. PPTX gives each page its own slide, and every line of text and every bubble stays editable in PowerPoint — except a mirrored text box or table, which PowerPoint cannot mirror, so it comes across as a picture of itself.

DOCX is a Word document you can keep editing. Your text becomes real Word paragraphs in the Document's fonts, and each page starts where it starts in the editor. Every item sits on its own page where you put it, and text wraps around the items you've set it to wrap around. Because Word lays out text itself, a line can break a little differently from the editor — most often in a justified paragraph, where Word spaces the words its own way — and a font size between Word's half-point steps is rounded to the nearest one. A few things have no Word equivalent, so they come across as a picture of themselves: a mirrored or see-through item, a curved or tailed bubble, outlined text, and a turned or rounded table. In a header, a footer and the page's own text, capitals carry across and an outline does not — Word has no outline for words that flow.

Emoji inside text. Emoji you type inside a text box or the page's text show in your computer's own emoji style in PPTX and DOCX. An emoji placed on the page as an item comes across exactly.

Markdown and Text keep your words. Both keep every word the editor shows, in reading order: the header's words first, then the text on the page, then each page's items from top to bottom, then the footer's words. Lists keep their bullets and numbers, code keeps its lines, and tables keep their rows and cells. Markdown also keeps headings, bold, italic, strikethrough, underline, highlight, code, links and citations, and embeds your pictures with their captions; Text keeps a picture's caption. Neither format has pages, positions, alignment, colours, sizes or fonts, so those don't come across, and a header or footer slot that shows a page number is left out.

Undo and Redo

While editing, use Undo to step back through your recent changes and Redo to step forward again. This lets you experiment freely — try a formatting change or rewrite, and roll it back instantly if you don't like it.

• One history for the whole document. Typing in the body, in a text box or in the header and footer, a formatting change, and moving, resizing or drawing a box are all steps in the same list, undone in the order you made them.
• Undo takes you to the change. The cursor goes back into the box where the text changed and the page scrolls to it; undoing a header or footer change opens it again.
• Typing undoes a run at a time. A burst of typing is one step, so a pause or moving to another place starts the next one.
• A drag is one step. Dragging a slider such as the font size is one step, and so is every colour you try for the words — in the page's text, a text box, or the header and footer — while that colour picker stays open, so one Undo puts back what you had before you started.
• A 3D model is part of the same steps. While a 3D model's 3D or Transform tab is open, Undo and Redo take your changes back newest first, to the model and to the rest of the page alike; once you leave those tabs, one Undo takes back the whole change to the model, even while its picture is still being redrawn.
• The same step everywhere. The Undo and Redo buttons, Cmd+Z and Shift+Cmd+Z (Ctrl+Z and Ctrl+Shift+Z on Windows), and on a touch screen a two-finger tap to undo or a three-finger tap to redo all take the same step.