The nine flowss studios are rooms of one building, and your work can move between them. A Mermaid flowchart can become an editable Weave canvas, a FlowScript document in the Lab, a BPMN process, a spaced-repetition course in Study or an evidence scaffold in Evidence, and a rough Sketch board can become a diagram anywhere. This page explains every way to move work, how flowss decides what each move does (an exact conversion, a cached view, or an AI rewrite), what each studio can send and receive, what survives the trip, what each move costs, and how flowss keeps track of where your work came from.
At a glance
| You want to | Use |
|---|---|
| Open a Library document in another studio | The card's … menu → Open in → a studio |
| See exactly what a conversion will do first, or keep the result as a separate document | The card's … menu → Open as… |
| Send what you are working on to another studio | Send to… in the studio (in the Share panel's Send section, or the studio's own toolbar or export menu) |
| Hand it on from anywhere, in one click | The Flow Dock (⌘J) → Take it further |
| Turn a diagram into a course | Study these N or Turn into N lessons in the Library, Study it in the Flow Dock, or Send to… → Study |
| Turn a diagram into an evidence investigation | Research it in the Flow Dock, Send to… → Evidence, or Open in → Evidence |
| Pick up where you left off | The Flow Dock's Continue, the Library's Recently open rail, or ⌘K |
| See how your work has travelled | Journey, in the Flow Dock and above the Library board |
Which studios send and receive
Not every studio does both. A studio "receives" when a diagram can be delivered into it, and "sends" when it can hand its own work on.
| Studio | Receives | What it receives as | Sends | What it sends |
|---|---|---|---|---|
| Studio | Yes | Any engine, exactly as it is. The Studio renders every engine | Yes | The open figure's source |
| Study | Yes | Any diagram, as the notes for a new course | No | Study's courses stay in Study |
| Weave | Yes | An editable Weave canvas | Yes | The canvas |
| Lab | Yes | A FlowScript document | Yes | The FlowScript document |
| BPMN | Yes | A BPMN 2.0 process model | Yes | The process model |
| Science | Yes | An editable node-and-connection canvas | Yes | The canvas |
| Figure | No | Figure works from a dataset or a pasted SVG | Yes | The composed figure as an SVG, or a selected panel (with the source that drew it, when the panel came from data) |
| Sketch | No | Sketch is a drawing surface | Yes | The board, read as a Mermaid diagram from your shapes and connectors |
| Evidence | Yes | Any diagram, turned into an evidence scaffold to research | Yes, once a review has a result | The review as a Mermaid diagram: the PRISMA flow in the screening pipeline, the causal graph in the causal lens, otherwise the concept map |
The Studio is the universal destination: anything that can be rendered opens there unchanged.
Evidence is the one studio whose Send to… is not counted as a source by the Flow Dock: its diagram is drawn from the review at the moment you send, so use the Send to… in Evidence's Share panel. When a PRISMA flow is sent to the Lab, the Lab version is written straight from the counts with no AI. Sent anywhere else, an AI conversion is checked first, and refused with "The conversion changed the PRISMA counts, so it was not sent — open the flow in Studio instead, which renders it as it is." if any count did not survive.
How flowss decides what a move does
Whenever work moves to a studio that does not natively read it, flowss picks the cheapest way that works, in this order:
| Way | Badge in Open as… | What happens | Costs |
|---|---|---|---|
| Opens directly | opens directly | The studio already reads this document as it is | Nothing |
| Cached view | ready | You opened it there before and the document has not changed since, so the earlier conversion is reused | Nothing |
| Layout | exact, no AI | A diagram with nodes and connections is laid out as an editable canvas in Weave or Science. Every node and label is kept | Nothing |
| Transpile | exact, no AI | A built-in translator rewrites the source into the target's language | Nothing |
| Build | builds | Study builds a course from the document, or Evidence extracts its concepts and relations into a scaffold | Nothing |
| AI rewrite | AI | The flowss Library Agent rewrites the diagram in the target studio's language, keeping every node, connection, label and grouping it can express | One standard AI call |
Where each target's conversions usually land:
| Target | From diagram source (Mermaid, PlantUML, Graphviz, D2 and so on) | From a Weave or Science canvas |
|---|---|---|
| Studio | Opens directly (BPMN models too) | Not offered: a canvas has no source for the Studio to render |
| Weave, Science | Layout, exact, no AI, when the diagram has nodes and connections; otherwise an AI rewrite into a canvas | Opens only in the canvas studio that made it |
| Lab | AI rewrite into FlowScript, unless it is already FlowScript | Not offered |
| BPMN | AI rewrite into BPMN, unless it is already BPMN | Not offered |
| Study | Build | Build |
| Evidence | Build | Listed in Open in, but fails from the Library (see the warning below). Use the canvas studio's Send to… → Evidence |
A Sketch board is ink rather than source, so it must become a diagram first: use Sketch's Send to…, which reads the board as a diagram, or recognise the board in Sketch.
Warning: The Library's Open in list offers Evidence for every document, including Weave and Science canvases and Sketch boards. Those usually hold no diagram source (only an imported Weave file does), so opening one in Evidence from the Library fails with "Missing 'code'". Open the document in its own studio and use Send to… → Evidence there instead.
AI rewrites need AI on your plan: your own key from Starter, or hosted AI points from Plus. Free has no AI, so on Free only the exact and built moves are available. See AI points and limits and Bring your own AI key.
Opening a Library document in another studio
- In the Library, open the document card's … menu.
- Under Open in, pick a studio. The list holds every other studio that can show this document, cheapest first. A hint reads ready when a converted view is already saved, and AI when opening there needs an AI rewrite. If no other studio can show it, the list says "Nowhere else yet".
- flowss converts the document if it needs to (an AI rewrite shows "The flowss Library Agent is rewriting this as *engine*…"), hands it to the studio and opens it.
What happens to the document:
- The converted form is remembered on the document, so opening it there again is instant and free.
- Edits you make in that studio update the remembered form, not the original. The original's source stays exactly as it was.
- If the original changes later, the next open rebuilds the converted form from the new original.
To keep a converted version as a document in its own right, with its own title and history, use Save copy in Open as… below.
Opening the document in its own home studio (by clicking the card's thumbnail) is always a direct open, and your edits there are written back to the document itself.
The Open as… dialog
Open as… is the careful version of Open in. It shows what each move will do before it does it, and lets you keep the result.
- Open the card's … menu and choose Open as… (hint "details"). In the list view it is in the row's … menu.
- The dialog, titled **Open "*title*" as…**, explains: "Opening as a view keeps this one document and remembers the derived form. Saving a copy creates a new document that knows where it came from."
- Each studio that can show the document has a row with its name, a badge (see the table above), and one sentence describing what will happen, for example "Lays the diagram out as an editable Weave canvas — exact, no AI." or "Asks the model to rewrite this as bpmn."
- Choose an action on a row:
| Button | What it does |
|---|---|
| Open | Opens the document in that studio as a view, exactly like Open in |
| Save copy | Converts the document and saves the result as a new Library document in that studio, titled "*title* (Studio)", for example "Checkout flow (Weave)". A toast says "Saved … — linked to …". The new card shows "from *original title*" |
Save copy appears only for real conversions (ready, exact, no AI and AI). It is not offered where the document already opens directly, or where the target builds something of its own (Study and Evidence).
If no studio can show the document, the dialog says "Nowhere else can show this document yet."
The sentences you may read in the dialog:
| Sentence | Meaning |
|---|---|
| "Opens directly in Studio." | No conversion needed |
| "Reopens the Studio view you already built." | A cached view is reused |
| "Lays the diagram out as an editable Studio canvas — exact, no AI." | Layout into Weave or Science |
| "Rewrites the source to engine — exact, no AI." | A built-in translator |
| "Asks the model to rewrite this as engine." | An AI rewrite |
| "Builds a course from this document — learn it with spaced retrieval." | Study |
| "Extracts the concepts and relations into an evidence graph." | Evidence |
Note: On Free, Save copy is refused in a studio that already holds 3 documents, before any AI call is made, so you are never charged for a copy you cannot keep.
The Send to… menu
Every studio that sends has a Send to… control ("Send this diagram to another studio"). In studios with the island layout it is in the Share panel's Send section; in Figure it is also in the export menu for the whole figure and on the toolbar for a selected panel. In the Studio, the ⋮ menu's Send to another studio… (also in ⌘K) opens the Flow Dock's Take it further chips as a quicker alternative. Send to… is disabled while there is nothing to send.
The menu has two parts.
Send to another studio lists every studio that can receive this diagram, except the one you are in. Each row shows the studio's name, its one-line description, and a convert chip when the diagram is not in that studio's own language and must be converted. A conversion uses a built-in translator when one can read your diagram (for example a Mermaid flowchart into a Weave canvas), at no cost; otherwise it is an AI rewrite. Rows:
| Studio | Description shown |
|---|---|
| Studio | "The universal canvas — renders every engine." |
| Study | "Interactive knowledge maps, model-building and spaced retrieval." |
| Weave | "Drag-and-drop diagramming canvas." |
| Lab | "FlowScript — typed, composable diagram code." |
| BPMN | "Business-process modelling (BPMN 2.0)." |
| Science | "Biology & science figure composer." |
| Evidence | "Turn this diagram into an evidence scaffold to research." |
Not for diagrams lists the studios that cannot receive a diagram (Figure and Sketch), dimmed, with the reason, for example "Figure works from its own material, so a diagram can't be delivered there." A footnote explains: "“convert” diagrams are rewritten by AI into the target studio's native format; dimmed studios work from their own material rather than a delivered diagram."
What happens when you pick a studio
| Target | What happens |
|---|---|
| A studio that renders this engine (always the Studio) | Delivered at once. A toast says "Sent to *Studio*" and the studio opens |
| A studio that needs a conversion (convert) | flowss converts the diagram, then shows a preview: **Open in *Studio***, "converted to *engine*", a small rendering of the result, and any notes about what could not be carried over. Click **Open in *Studio* to go, or Cancel (or Back**) to stay |
| Study | Delivered as course notes. "Sent to Study — Study will build a course from it" |
| Evidence | Builds an evidence scaffold from the diagram and opens Evidence. "Sent to Evidence" |
Where the diagram lands in the target:
| Target | Lands as |
|---|---|
| Studio | A new figure |
| Lab | A new FlowScript document |
| BPMN | A new tab in the BPMN workspace |
| Weave, Science | The studio's canvas. These studios hold one live canvas, so the delivery replaces what is on it |
| Study | Notes in Study's course creator |
| Evidence | A new scaffold to research |
The target studio is detached from the Library document it had open before the delivery is written, so the delivery is never saved into that document.
Tip: Weave and Science hold one canvas at a time. If what is on the target canvas matters to you, save it to the Vault from the Flow Dock before you send something there. See Version history and undo.
If a conversion fails, the reason stays visible at the bottom of the menu, for example "Couldn't convert to bpmn. Try opening in Studio instead." or the AI service's own message.
What each studio sends
- Studio, Weave, Lab, BPMN, Science send the document you have open, in their own language.
- Sketch reads your board as a Mermaid diagram at the moment you send, from its shapes, text and connectors. If the board has no diagram in it yet you see "This board can't be sent yet — reason." For a fuller conversion, use Sketch's own recognition first. See Sketch.
- Figure sends the composed figure as an SVG from its export menu, or one selected panel from the panel toolbar. A panel made from data carries the chart source that drew it, so it arrives as an editable chart. See Figure.
- Evidence sends the review it has on screen as a Mermaid diagram (the PRISMA flow, causal graph or concept map), from the Send to… in its Share panel, once there is a result. See Evidence.
The Flow Dock's Take it further
The Flow Dock (⌘J, or its floating pill) holds a Take it further section while you are in a studio with something to hand on. It offers up to three one-click destinations, ranked by what you tend to do next (a studio you just left is ranked lower). When there is no better suggestion, the dock falls back to Research it, Study it and To Studio:
| Chip | Tooltip | Does |
|---|---|---|
| Research it | "Build an evidence scaffold from this diagram" | Sends to Evidence |
| Study it | "Turn this diagram into a course" | Sends to Study |
| **To *Studio*, for example To Weave** | "Open this in *Studio*" | Sends to that studio, converting first when needed |
A chip runs the same delivery as Send to…, without the preview step. A rewrite shows "Converting to *engine*…" first, and if a built-in translator could not carry everything, a second toast names what was dropped.
When the studio's document cannot travel through the dock, the section shows why instead of the chips:
- "*Studio* converts its document before it can travel — use *Studio*'s own Send." (Sketch and Figure, whose Send to… does the reading or composing first)
- "This canvas can't be read from storage, so it can't be handed on. Reload Studio and try again." (a Weave or Science canvas whose saved copy is damaged)
In Study and Evidence, and in any studio with nothing open yet, the Take it further section simply does not appear. Evidence can still send from its own Send to….
Studying a document
Any diagram or canvas document with at least two ideas (nodes) that Study can read can become a course in Study. Study is part of Starter and up.
| Where | Control |
|---|---|
| Library card | Turn into N lessons at the foot of the card, or … → Study these N |
| Library list row | … → Study this |
| Flow Dock | Study it |
| Any studio | Send to… → Study |
From the Library, a course is built from the document's concepts and connections, about one lesson for every four concepts. The course is linked to the document:
- Studying the same document again reopens the same course, with your progress intact. It never makes a second, empty copy.
- If the document has changed since, the course content is refreshed to match, unless you have edited the course yourself, in which case your version is left exactly as you have it.
A document with fewer than two readable ideas shows Not studyable yet in its card menu, or the message "This one has no structure to learn from yet — connect a few ideas first." when you choose Study this in the list view.
Sending from a studio (Send to… → Study, or Study it) hands the diagram to Study's course creator as notes, which you then turn into a course there. See Study.
Researching a document in Evidence
Sending a diagram to Evidence turns its concepts and relations into a plain-language scaffold (a research question, a title and source text) that Evidence can investigate against real papers. Diagrams that cannot be read as nodes and connections are sent as their source text. This step uses no AI points, but it counts toward your daily rendering allowance, and the diagram source can be at most 100 KB. Evidence itself requires Starter and up. See Evidence.
Lineage: where your work came from
Every Library document keeps an account of how it came to exist: created, duplicated, converted, imported, and so on. You see it in three places:
- "from …" on Library cards. A document derived from another shows "from *parent title*" with a branch icon. Hover it for "Derived from …". Duplicates, Save copy results and imports all carry it.
- Search. The notes in a document's history, such as "mermaid → lucidflow" or "imported roadmap.mmd (mermaid)", are searchable.
- Journey. Each time you send work from one studio to another, the move is recorded. The Library's Journey band shows each route with a count, such as "Sketch → Studio ×3" (hover for "3 handoffs, last 2h ago"). The Flow Dock's Journey section lists recent moves as "*From* → To · title · *time*".
Moves started from the Library, the Flow console or the command palette are not recorded in Journey, because they do not start in a studio.
Picking up where you left off
| Where | What it shows |
|---|---|
| Flow Dock → Continue | Your latest thread through the studios: each recent document with its studio and time. Click to go back to it |
| Library → Continue band | The same thread, above the board |
| Library → Recently open | The documents each studio has open right now |
⌘K | Your documents, searchable from any page; picking one opens it in its home studio |
| The Flow console | The full-page view of where your work is and the next move worth making. See /docs/flow |
While you are signed in, Continue and Journey follow you to your other devices within moments.
What survives a move
| Move | What carries | What does not |
|---|---|---|
| Any diagram to the Studio | Everything | Nothing is lost |
| A node-and-connection diagram to Weave or Science (layout) | Every node, connection and label | The source engine's own styling; you arrange the canvas from there |
| Any diagram to the Lab or BPMN (AI rewrite) | Every node, connection, label and grouping the target language can express | Anything the target cannot express. The preview's notes say what |
| A built-in translator (for example Mermaid flowchart to D2) | What the target can express; a toast names anything dropped | Constructs with no equivalent |
| A Weave canvas to Science, or the reverse | Not offered | The two canvases use different shapes (Science draws illustrated icons, Weave draws flowchart shapes), so flowss will not flatten one into the other. Open a canvas in the studio that made it |
| A Sketch board to anywhere | Shapes, text and connectors read as a diagram | Freehand strokes that are not shapes |
| Anything to Study | Concepts and connections, as a course | Layout and styling |
| Anything to Evidence | Concepts and relations, as a scaffold | Layout and styling |
The original is never changed by a move. Conversions write to the target studio or to a new document, never back over the source.
Plans and access
- Studio access. Free opens the Studio and Sketch. Weave, Lab, BPMN, Science, Figure, Study and Evidence start on Starter; sending or opening work in one of them on Free takes you to the pricing page for that studio. See Plans and what they include.
- AI rewrites. One standard AI call each, on your own key (Starter and up) or hosted AI points (Plus and up). Exact conversions, cached views, Study and Evidence moves cost nothing.
- Document limits on Free. Save copy and anything else that adds a document is refused in a studio that already holds 3 documents.
Tips
- Prefer exact moves. A diagram with clear nodes and connections lays out in Weave or Science with no AI and no loss. Check the badge in Open as… before you choose.
- Save a copy when the conversion is the point. If you want to keep developing the Weave version separately from the Mermaid original, use Save copy; a plain Open keeps it as a view that is rebuilt if the original changes.
- Preview before you leave. Send to… shows a rendering of any AI conversion and its notes before you commit; Cancel keeps you where you are.
- Study the same document again to refresh a course. It reopens the same course with your progress, and updates the content unless you have edited the course.
- Use the Studio when in doubt. Every diagram opens there exactly as it is.
Limits and known constraints
- Figure and Sketch cannot receive a diagram. Study cannot send one, and Evidence sends only from its own Send to…, once a review has a result.
- From the Library, Open in → Evidence fails for Weave and Science canvases and Sketch boards, which hold no diagram source.
- Weave and Science canvases cannot be opened in each other.
- Sending to Weave or Science replaces the canvas that studio has open. Checkpoint it in the Vault first if you need it.
- Edits made in a converted view are kept with that view and are replaced if the original changes and the view is rebuilt. Use Save copy to keep them independently.
- AI rewrites need AI on your plan and keep only what the target language can express.
- Sketch sends what it can read as a diagram from your board; freehand strokes that are not shapes do not travel.
- Journey records moves that start in a studio, not moves from the Library, the Flow console or the command palette.
- Moving work into a studio your plan does not include takes you to pricing rather than opening it.
Troubleshooting
| Message or symptom | What it means | What to do |
|---|---|---|
| A studio is missing from Open in | That studio cannot show this kind of document | Use Open as… to see what each studio can do, or open it in the Studio |
| "Nowhere else yet" / "Nowhere else can show this document yet." | No other studio can show this document | Open it in its home studio |
| "Recognise the sketch first to turn it into a diagram." | A Sketch board's ink has no source to convert | Recognise the board in Sketch, or use Sketch's Send to… |
| "*Studio* can't open a canvas document — open it in Weave or Science instead." | A canvas has no diagram source for that studio to read | Open it in the canvas studio that made it |
| "*Studio* doesn't take documents — it works from your own material." / "*Studio* works from its own material, so a diagram can't be delivered there." | Figure and Sketch do not receive diagrams | Choose another studio |
| "The flowss Library Agent is rewriting this as *engine*…" then an error | The flowss Library Agent's rewrite failed, for example because you are out of AI points or have no AI on your plan | Check AI points and limits, or choose an exact move or the Studio |
| "The flowss Library Agent's rewrite didn't come back as a Studio canvas — nothing was staged." | Its answer could not be read as a canvas | Try again, or open the original in the Studio |
| "Couldn't convert to engine. Try opening in Studio instead." | The conversion returned nothing usable | Open in the Studio, which renders the original as it is |
| "Couldn't prepare this for Studio." | An exact conversion produced nothing usable for that studio | Try Open as… to see the other routes, or open it in the Studio |
| "Couldn't turn this into a Studio document." | Save copy could not build a document from the conversion | Use Open instead, or try another studio |
| "Couldn't stage the diagram (storage full)." / "Couldn't stage the document (browser storage is full)." | The browser would not store the delivery | Free space from the Library's Storage health panel, then try again |
| "Couldn't build an Evidence scaffold from this diagram." | The diagram could not be turned into a scaffold | Check the diagram renders, or paste its text into Evidence directly |
| "Missing 'code'" | Open in → Evidence from the Library on a Weave or Science canvas or a Sketch board, which has no diagram source | Open it in its own studio and use Send to… → Evidence |
| "Code is too long (>100 KB)" | The diagram sent to Evidence is over 100 KB of source | Send a smaller diagram, or split it |
| "Rate limit reached (*N*/day)." | Today's allowance for diagram processing is used up | Try again tomorrow |
| "Couldn't stage the Evidence scaffold (storage full)." | The browser would not store the scaffold | Free space from the Library's Storage health panel, then try again |
| "The conversion changed the PRISMA counts, so it was not sent …" | Sending Evidence's PRISMA flow, the conversion lost or altered a count | Send it to the Lab (written exactly from the counts) or to the Studio (rendered as it is) |
| "This board can't be sent yet — …" | Sketch found no diagram on the board | Add shapes and connectors, or recognise the board first |
| "Still reading the board — try that again in a moment" | Sketch is still loading its reader | Wait a second and send again |
| "This one has no structure to learn from yet — connect a few ideas first." | Too few connected ideas to build a course | Add nodes and connections |
| "Free accounts keep 3 figures in a studio, and this one is full. …" | Save copy into a full studio on Free | Make room in that studio, or upgrade |
