Profusia

How can we help?

Ask a question about using Profusia — connecting your AI, publishing, sharing, sites, live data. Everything below is also written out, if you'd rather read.

Answers come from the help topics below, written by an AI. It can be wrong — if something matters, check the topic it came from or email us.

Help topics

What Profusia is

Profusia is a place for AI-generated documents to live. You connect your AI assistant (Claude, ChatGPT, or anything else that speaks MCP) to Profusia, and it can publish files straight into your workspace — web pages, dashboards, Markdown, code, images, PDFs, Office documents. Published documents get a private URL you can share, they keep a full version history, and they can be arranged into a site with navigation. Pages can be kept up to date automatically with live data. The short version: your AI makes something, and instead of a file in a chat window you get a real link that stays current.

This answer as its own page

Connecting Claude to Profusia

Profusia connects to Claude as a custom connector. In Claude, go to Settings → Connectors → Add custom connector, and give it your workspace's MCP URL, which is your Profusia address followed by /mcp — for example https://yourteam.profusia.com/mcp. Claude will walk you through signing in (Profusia uses standard OAuth). Once it is connected, just ask Claude to publish something: "publish this as a page in Profusia" works. There is also a static access-key option if you prefer, which your workspace owner can issue.

This answer as its own page

Connecting ChatGPT to Profusia

ChatGPT works too, and there is nothing extra to set up on Profusia's side — the connection uses the same open standard (MCP). In ChatGPT, enable developer mode and add a custom connector pointing at your Profusia MCP URL (your address followed by /mcp). When it asks for authentication, the API-key option takes your workspace access key. One honest caveat that is OpenAI's rule, not ours: as of late 2026, ChatGPT's developer-mode connectors with full read-and-write access are available on ChatGPT's Business, Enterprise and Edu workspace plans — not on the free or Plus tiers. Claude users on every plan can add a custom connector. Everything works the same as it does in Claude: publishing, updating, share links, live data. ChatGPT's scheduled tasks can also refresh live data on a schedule, the same way Claude's Routines can.

This answer as its own page

Publishing a document

Ask your assistant to publish. For example: "publish this report to Profusia" or "put this dashboard in Profusia under the Client Work folder". Your assistant will write the file and send it over. You get back a link, and Profusia tells the assistant where the file landed so it can tell you the folder too. You can publish HTML pages, JSX (which Profusia compiles in the browser — no build step), Markdown, code files, images, PDFs, and Office documents. Anything Profusia can render becomes a live page; anything it can't becomes a clean download. Word, PowerPoint and Excel files are searchable by what they say, not just their name — Profusia reads the words inside them when they are published, so they show up in search and can be asked about like any other document. A PDF is still matched by name only; Profusia says so rather than pretending to have read one it hasn't.

This answer as its own page

Templates — a status page, weekly numbers, or a quote in one ask

Profusia ships ready-made page templates your assistant fills in: a client status page, a weekly numbers report, and a quote. Say something like "publish a client status page from Profusia's template" — your assistant asks a few questions (the project, the milestones, what's next) and publishes the finished page. The status page and weekly numbers are live: they come with a small dataset behind them, so the weekly update is a data refresh (update_dataset) rather than a new document — every link you already shared shows the new numbers. The quote is deliberately a snapshot: its prices never change after you send it. You can preview every template, filled with sample data, at profusia.com/templates. (list_templates / publish_from_template)

This answer as its own page

Writing a document yourself, in the browser

You do not need an assistant for everything. On any site, the Add button (and the + beside any space in the navigation) offers "Write a document". It is a normal rich-text editor: headings look like headings, bold looks bold, and you never see Markdown syntax. Paste from Word or a web page and the formatting that Profusia can keep is kept. You can drop pictures straight in. What gets stored is Markdown, which is deliberate — it is what lets your assistant pick the document straight back up later and revise it. Occasionally a document opens showing Markdown source instead, with a line explaining why. That happens when it contains something the visual editor cannot hold without changing it — a table, or a nested list. Profusia checks each document before opening it and shows the source rather than risk quietly altering your content. There is a Markdown / Rich text switch in the toolbar either way. To edit a document that already exists, open it and use Edit.

This answer as its own page

Uploading a file you already have

If the file already exists — on your computer or phone, exported from another app, attached to a chat — don't make your assistant retype it. Ask for an upload link. Your assistant can call get_upload_link, which gives you a private drag-and-drop page. Drop the file in and it lands in your workspace like anything else. That works for any file type, and it is much faster than streaming a large file through a conversation. You can drop a whole folder in too, or choose one with the folder picker — its structure is kept, and every file inside lands in the matching place in your workspace rather than being flattened onto one level.

This answer as its own page

Folders versus Projects

These do different jobs, and it is worth getting the distinction: A FOLDER is where a file lives — the filing cabinet. One file, one folder. A PROJECT (some workspaces call them Initiatives, Campaigns, or something else — you choose the word) is what a file is FOR. Projects can nest, a file can be in one, and each project carries a description and a set of goals. The reason projects matter: those goals get read back to your assistant whenever it publishes into that project or lists what is in it. So a session next month picks up what the project is trying to achieve without you re-explaining it.

This answer as its own page

Sites and spaces

A site is how your documents are presented to a reader — a navigable place rather than a pile of links. A site is made of SPACES, which nest up to six levels deep. Each space gets an overview page generated from what is actually in it: real counts, what changed recently, cards for the spaces inside it, and live previews of the documents. You can build a whole site in one call from a template that mirrors your existing folders, then refine it — either by asking your assistant, or by hand in the visual editor. Sites are live. There is no publish button. What controls who can see a site is simply whether a link to it exists.

This answer as its own page

Changing how a site looks

Two separate settings, and knowing which is which saves a lot of confusion. NAVIGATION belongs to the whole site: Sidebar (every space down the left), Top bar (a menu bar with drop-downs and a second row for wherever you are), or Minimal (just the wordmark, with Contents on demand). It is one choice for every page, so navigation does not change as you move around. LAYOUT belongs to each space — how that one page displays what is inside it. The choices are Showcase, File list, Dashboard, Story, Handbook and Shelf. A space inherits its layout from its parent unless you set one; Showcase is the default at the top of a site and File list inside it. Both are in the same menu on the space itself, clearly separated. You can also ask your assistant to set a layout (set_space_shell). To preview without committing, add ?shell=shelf to the URL for a layout, or ?nav=top for navigation.

This answer as its own page

Comments and feedback

Every document has a comment thread behind a small toggle, and every space has its own discussion for talking about the place rather than a single file. A comment on a document can be PINNED to a specific passage or image: select the text, and your comment points at exactly that. The highlight is drawn over the document — the document itself is never modified. Comments are for signed-in members of the workspace, any role. Someone holding a share link is anonymous, so they see the document but not the discussion. The useful part: your assistant can read the comments (read_comments) and revise the document based on them, then post a note saying what it changed. Pinned comments come through with the passage they point at, so "this number is wrong" is actionable. If you are hunting for a comment you remember, there is a Comment Center that shows every comment across the workspace, filterable.

This answer as its own page

Versions and undoing changes

Every time a document is updated, Profusia keeps the previous version. Nothing is overwritten. To go back, ask your assistant to roll the file back (rollback_file) to an earlier version, or do it yourself: open the document in Docs, and its panel has a Version history section listing every version with its date, who made it and any note they left. Press Restore — it asks you to confirm with a second press, then makes that version current. Either way, restoring never deletes anything: the version you were on stays right there in the list, which is what makes the whole thing undoable. The share link never changes, so anyone holding it just sees the restored content.

This answer as its own page

A document's own facts — owner, type, and review date

Every document can carry three facts of its own: who owns it, what kind of thing it is in your organization's own words (Policy, Proposal, Contract — whatever your workspace calls these), and when somebody should check it is still true. Ask your assistant to set them (set_document_facts — it does not cut a new version, unlike publish_file and update_file, which take the same three alongside the content), or set them yourself in the document panel in Docs, under "About this document". The box remembers whatever your workspace has already typed and offers it back, so typing consistently is all it takes to get a controlled list — there is no separate screen to manage one. Owner is a label. It grants nobody access and hides nothing from anybody — the panel and every tool response say so, because the one wrong idea this feature could create is "I set the owner, so only they can see it". A review date is kept exactly as you typed it — "end of Q3" is a real answer. Only a value Profusia can actually read as a date counts toward "overdue" on Needs attention, and that finding says plainly how many review dates it could not read rather than dropping them silently. Search matches owner and type too, and tells you which one matched.

This answer as its own page

Following a document

If you want to hear about it the next time a document changes, follow it. There is a Follow button on the document panel in Docs, next to Pin. Following is for signed-in members, not for an assistant over MCP — it is one person saying "tell me about this one", and a workspace access key has nobody to tell. It grants nothing in either direction: you can only follow a document you can already see, and following it never widens what you can see. Nobody else can see who follows what — recording that would be surveillance, not a feature. When a document you follow gets a new version, Profusia notifies you (there is no email yet — see "Not available yet"), and the Docs module also shows a line naming the documents you follow that moved this week, so you do not have to go looking.

This answer as its own page

Where a document stands, and who is holding it

A document can carry a state — draft, in review, approved, or superseded — so a reader can tell at a glance whether to rely on it. Ask your assistant to set one (set_document_state), or use the State field in the document panel. The four words are fixed, the same in every workspace: a state is a claim other people rely on, not an internal process each team should get to rename. And it is a LABEL — it gates nothing. A document marked "draft" is visible to exactly the same people who could already see it; the state just tells them what they are looking at. Every surface that shows it says so. Checking a document out is different: it really does refuse. Ask your assistant to check one out (check_out_document) before you start rewriting it, and nobody else can publish a new version — update_file, publish_file, rollback_file — while you hold it; they are told who does. Reading, renaming, moving and commenting are all unaffected. Release it (check_out_document again, or the Release button on the document panel) as soon as you are done — a forgotten check-out is worse than none, which is why the panel always says how long one has been held. A check-out belongs to a person, so it only works when your assistant is signed in as an account (an OAuth connection) — a workspace access key has nobody to be, and is told so plainly rather than silently taking a lock nobody can release. An admin can always force one open if its holder is gone.

This answer as its own page

Deleting, renaming and moving things

Deleting is reversible. When you delete a file, folder, site, project or dataset it goes to the workspace trash — it stops appearing in listings, and its share links stop working immediately, but it is not gone. You can see what is in the trash (list_trash) and restore it (restore_item). If you delete a folder, restoring it brings back what was inside as a group. Permanently erasing something is a separate, deliberate step (purge_trash) that you have to confirm. Nothing is ever swept away automatically. Renaming and moving work on anything: files, folders, sites, spaces, projects.

This answer as its own page

Live data — pages that stay current

A document can be backed by a small dataset so its numbers stay fresh without republishing the whole page. The flow: your assistant creates a dataset (create_dataset), binds it to the page (bind_file), and from then on updating the data (update_dataset) updates what the page shows. The best part is scheduling it. Claude can set up a Routine — or ChatGPT a scheduled task — that fetches whatever the numbers come from and calls update_dataset on a schedule. Your dashboard then updates itself every morning without anyone opening it. The Live module (in the switcher at the top of every page) lists every table and what it is keeping current. Each table has a page of its own showing its shape, its freshness and everything that reads it — and the rows are an editable grid, so a person can correct a number by hand without an assistant. Members with a writer role can also see Connections, the wiring between tables, pages and plans.

This answer as its own page

Putting a Profusia document inside another tool

You can embed a Profusia document into SharePoint, Confluence, Notion, Teams, or an intranet page. Ask your assistant for embed code (get_embed_code). It will ask where you are putting it rather than asking for technical details, and gives you something you can paste. The embedded view shows the document plus a small strip saying what it is, how current it is, and how to get back to the full version. Pasting a plain link into Slack, Teams or Notion also produces a proper preview automatically. Embed access is a separate permission from a share link: forwarding a share link does not let someone embed the document. You can revoke an embed at any time (revoke_embed).

This answer as its own page

Taking a document back to your AI

Every document carries one button at the top right — "Take to AI" — with everything that takes it somewhere else inside. Talk about it: opens Claude or ChatGPT with a prompt about the document already written. The app the document was originally published from is listed first, and if that app has a connector to your workspace the prompt carries the document's id, so your assistant can pull the actual document rather than guess from a description. Work on it in code: opens Claude Code or ChatGPT Codex, with the GitHub repository the document was written in already selected. This shows only for signed-in members of the workspace, and only when your assistant recorded a repository when it published — useful for design notes, runbooks and knowledge-base articles that sit alongside a codebase. You can pick a different repository before you go. An admin can switch these options off for the whole workspace, and anyone can hide them just for themselves from the same menu. Any assistant: a Copy button that gives you the same prompt to paste into any other AI tool. Next to the Take to AI button there is also a small open-in-new-tab icon: it opens the document on its own, outside the site, ready to copy or print.

This answer as its own page

Keeping track of how a project changed direction

Profusia can record the moments a project's direction moved — a branch, a pivot, something abandoned — with the reason and links to what it replaced or revised. Your assistant records these as they happen (record_turn), because it is in the conversation where the decision gets made. Publishing a document is not one of these moments; a real change of direction is. Later, get_course reads the whole thing back: what the project is doing now, what was tried and dropped, and DRIFT — the documents last written before the direction changed, which may still assume the old plan. If a project has no recorded history, Profusia says so rather than inventing one from publish dates.

This answer as its own page

Putting your company's branding on it

In the workspace settings (the /console address) there is a Brand screen, for owners and admins. You can set your company name, upload a logo, choose an accent color, write the line people read when they land on your address signed out, and add a footer line for documents you share. Once branding is on, your logo replaces the Profusia mark across your sites, on documents shared by link, and on your own sign-in page. A preview on that screen shows exactly what it will look like. Each site can still choose its own color palette on top of this — the brand is the company, the palette is the site.

This answer as its own page

Team members and roles

A workspace has four roles: owner, admin, member and viewer. Owners and admins can invite people and manage the workspace; members can create and edit; viewers can read and comment. To invite someone, ask your assistant (invite_member) or use the admin area. Registration is invitation-only — you cannot sign yourself up to somebody else's workspace. You can also create groups and use them to narrow who can see a particular document (set_file_audience), which is useful when most of a site is for everyone but a few documents are not. The same narrowing works on a whole space (set_space_audience) — "this whole area is HR-only" — and everything inside it inherits the restriction. One thing worth knowing: narrowing a space hides the space and its navigation, not the documents themselves, so a document somebody already has a direct link to still opens unless you narrow that document too.

This answer as its own page

Security and privacy

The main points: Documents are private by default and only reachable with a link you created. Every workspace is served from its own subdomain, so one workspace's content cannot be loaded from another's address even with a valid link. Hosted pages run under a strict security policy: a page can only fetch its own dataset, and Profusia never executes anything from a document on the server. Profusia stores no third-party credentials. It has no access to your email, your Google Drive, or your Jira, and it never asks for them. All tokens and keys are stored hashed, expire where appropriate, and can be revoked. There is a fuller public write-up at /trust, which is the page to send to a security reviewer.

This answer as its own page

Asking questions about your own documents

If the people running your Profusia have configured an AI provider, two extra things appear. "Ask" on a document answers questions about that one document. "Discuss" on a space answers questions about the work in that place — it reads the documents there, the description of what the space is for, the project's goals, and the record of how its direction changed, and it tells you which documents it actually read so you can check the answer. Neither can change anything. They read and answer; they cannot publish, edit, delete or share. Nothing you ask is stored. Close the panel and the conversation is gone. If you do not see these, AI is not switched on for your deployment. That is a setting for whoever runs Profusia, not something a workspace admin can turn on, and it is a perfectly normal state to be in — everything else works the same.

This answer as its own page

The working guide for your assistant

There is a short guide at /skill written for your AI rather than for you. Saving it into Claude or ChatGPT (as a skill, project instruction, or custom instruction) makes your assistant noticeably better at using Profusia — it stops asking you where things go and starts recording project history properly. It is a public page, so you can hand the link to a colleague without giving them access to anything.

This answer as its own page

Plans — tracking the work behind the documents

Plans is Profusia's work-tracking module, alongside Docs, Live and Audience in the switcher at the top of every page. A plan breaks a piece of work into cards — and each plan picks its own word for them: steps, features, phases, whatever your team says. A card carries a status from your workspace's own vocabulary (rename them in the workspace settings), can hold documents, a checklist, comments, and fields like Owner, Due, and Priority. The status shows on the documents themselves, everywhere they appear, so a reader knows a page is still exploratory before treating it as decided. One plan opens as a live canvas with seven views — List, Board, Timeline, Map, Flow (which draws what waits on what), Tree (the plan as a mind map, folded until you open a branch) and Abacus (every card as one bead, on a wire per branch, in the station its status is in) — and nothing reloads as you work. You choose which of them appear and in what order: the Views control on the view bar opens a sheet where you keep the ones you use, move them with the arrows, and hide the rest — and the first one in your order is where a plan opens. The Plans front page (Overview) answers two questions: how all the work is going, and what is on you — "My work" gathers every card across the workspace whose Owner names you. A card can repeat on a rhythm (weekly, monthly, quarterly…), a plan can be drafted from a plain-language brief and accepted card by card, and any card can be handed to your AI assistant with its documents attached. A plan can leave as Markdown or a spreadsheet, and a read-only link can show an outsider the work's state without the documents. Ask your assistant to "create a plan" to start one, or use the New plan button on the Plans page.

This answer as its own page

Chat — conversations in the workspace

Every conversation in Profusia is a channel. Some point at a subject — a document's comment thread, a space's discussion, a card's comments — and some are free-standing rooms your team opens by name. Members can talk in any of them; channels are listed in the Plans rail under "conversations", and documents and spaces carry their own threads behind a small toggle. @mentions resolve to members when you write them. Nothing is delivered anywhere — no emails, no push notifications. Unread marks inside Profusia are the whole mechanism, which is deliberate.

This answer as its own page

Needs attention — what is slipping between the modules

Profusia's modules each answer within their own lane. Nothing else catches what falls between them — a plan with no documents behind it, a live table nobody reads, a page quietly serving eight-week-old numbers, a document published onto no site at all. "Needs attention" is the members-only view that does. It is in the account menu on every module page (and on site pages), and it opens on a list of the actual things, grouped under your work, your live data and your documents, each one a link you can click straight into. A panel on the left lists every finding so you can jump between them. Everyone in the workspace sees it, and everyone sees the same things — it is not an administrator's screen. At the foot of that page, under "Analysis", is the Depth Map: the shape of your whole library in one picture — which areas hold the most and what is quietly aging. Nothing on it needs doing; it is there for proportion. Everything here is computed from real activity — nothing in any of it is estimated — and a young workspace is told there is not enough to draw yet rather than being shown an impressive picture of nothing.

This answer as its own page

Pinned — keeping the things that matter to hand

Anything you keep coming back to can be pinned, and everything you have pinned sits behind one control — the pin, at the top right of every module page. You can pin a document, a plan, a single card inside a plan, or a live table. They all land in the same short list, grouped by which part of the product they come from, each showing its current name and, where it has one, its current status. To pin something: right-click it in any list — or open the ⋯ menu on its row — and choose Pin. A plan has a Pin button on its own header, and a document or a card has one in its detail panel. Unpin from the same places, or from the ✕ on the row in the list. Two things worth knowing. Your pinned list is yours — nobody else in the workspace can see it, and there is no way to pin something for the team. And a pin points at the thing rather than copying it, so a renamed document shows its new name, and if something is deleted the list tells you it is gone instead of leaving you a broken link — restoring it from the trash brings the pin back.

This answer as its own page

Audience — who outside can see what

Audience is the fourth module in the switcher: every door to the outside world, in one room, with revoke where the row is. Pages lists every outward door — share links, embeds, portals, read-only plan links — sorted by how much each one reaches. People shows who has come through the named doors. Reach shows whether anyone is actually looking: opens are counted, but visitors are never identified. A PORTAL is how you share a living part of your site with a customer: one space (or the whole site), read-only and always current. It can be an open link, or a named door where each visitor enters their name — named visitors can be revoked one at a time. Ask your assistant to create one (create_portal), or review and revoke from the Audience module. There is also "See it as…": you can walk your own site exactly as a visitor or as one portal's audience sees it, so you can check what is reachable before you send the link. While you are seeing-as, nothing can be changed — it is strictly a viewing mode.

This answer as its own page

Finding your way — the Navigator

The small compass at the top right of every page opens the Navigator: one place that can find anything, teach you the product, and answer questions. Type, and it matches as you type — pages, spaces, documents (including the text inside them), plans, cards, live tables, and the product's own features. Press Enter and you are taken straight there; where it matched text inside a document, the passage is highlighted when you arrive. Guides are short click-through walks — what each module is for, how sharing works, what the AI can do — built from this help, with a "Show me" that takes you to the real page. If AI is switched on for your deployment, you can also just ask a question in your own words — about the product, or about your own workspace ("which plans mention the pricing work?"). It reads only what you can already see, tells you exactly which documents it read, and offers the places to go next. Nothing you ask is stored. If AI is not switched on, everything else still works — that is a normal state, not an error.

This answer as its own page

Exporting your data

Everything can leave. In the workspace settings (the console), the export takes your whole account: every document (every one you can see), plans as Markdown, datasets as JSON, site structures, your projects list, and the audit log as a spreadsheet — zipped in your browser, so the archive never sits on a server. One plan can also be exported on its own — Markdown or CSV — from the Share control on its canvas. And every document has always been downloadable from its own page.

This answer as its own page

Not available yet

Being straight about what Profusia does not do today: There is no published pricing and no self-serve signup — Profusia is in development and workspaces are set up by hand. Email is not wired up yet, so invitations are links you pass on yourself, and there is no "forgot password" email. A workspace owner resets a password. Single sign-on is built but switched off, pending configuration. Profusia does not connect OUT to other systems — it will not read your Jira, Salesforce, or Google Drive, and this is a deliberate design decision rather than a missing feature. Your AI assistant already has those connections; the intended pattern is that your assistant reads them and publishes the result into Profusia. There are no email or push notifications of any kind yet — mentions and unread marks live inside Profusia only.

This answer as its own page