Pages don't have to be reachable only from inside MojoPad.
Right-click any link — a page link or a web link — and choose Edit Link… to change its text and destination without unlinking and starting over. (Clicking a link always navigates; the right-click menu is the editing door.) The same menu offers Open in New Tab / Open Beside for page links; for web links Import This Page from the link's site, Open Your Saved Copy when you saved that page in the wiki, Open in Browser and Copy URL; and Remove Link, which keeps the text.
Import This Page is for reading something and wanting what it cites. It captures what the link points at and files it in your wiki as a page — trimmed to the article, searched and linked like anything else — and it does not take you there. You are in the middle of something, and a page you are reading often cites more than one source worth keeping, so it stays out of your way and puts an Open it button at the foot of the window for when you are ready. It asks once how to keep the clip, cleaned up or as the page looked; answer that in Settings ▸ Web Clipper and it never asks again. It is an import, as its name says, so the page goes to your import folder (Settings ▸ Files & Backups ▸ Importing files) and not to your clip folder. On a link to a paper on PubMed it brings in the paper's record and abstract instead of the web page (see A PubMed address brings in the record, below).
You saved the article, and a link to it still went to the web. Not any more. Three cases this is for:
How it works. Click a link to the address of a page you saved in this wiki — with
Import This Page, File ▸ Import URL…, a board's Add web page here…, Explore's
Save to wiki or a clip of the whole page from your browser —
and your copy opens instead of the site, where a paper opens: in a tab of its own, beside your page, or in your
paper pane (see One pane for the papers you follow, in Writing a Long Paper). The status line
says so, with the day you saved it: Your saved copy, from October 3, 2026 — ⇱ opens
the original. The ⇱ and the site's name at the top of your copy open the page on
the web, and so does Open in Browser on the link's right-click. Holding ⌘ never
sends you to the web instead: it does to your copy what it does to any paper. With Open Beside the
page you are in chosen in Settings, it opens your copy here, over what you are reading; with
When a link opens a paper set to In the same pane as the last paper, it gives your copy a
pane of its own; otherwise a paper takes a tab of its own either way, and ⌘ changes nothing.
Such a link is followed by (saved), drawn the way (PDF) is: your page keeps the words you typed, and nothing extra appears in an export. A link that names a paper you keep as a PDF — by its DOI, its PubMed or its PubMed Central number — still opens the PDF first.
Which addresses match. The ones a person would call the same page: http or https, with or without www, with or without a slash at the end, with the tracking a newsletter adds, or with a heading or a quoted sentence after the #. A different page of the same site is a different page. A page you save from now on through an address that sends you on to another — a doi.org link, a short link, an address that moved — is found by both addresses. A page saved earlier is found only by the address it ended up at, because the one you clicked was not kept then. When two pages were saved from one address, the one you have marked, highlighted or written notes on opens, and otherwise the newer one. A page with its own password is never opened this way, locked or not, and neither is a page in Recently Deleted. Giving a page its own password also lets go of the link you saved it through, a doi.org or short link. The address it ended up at stays readable beside the password, as the page's name does and as it always has; a password on the whole wiki seals both. Take the password off later, and the page is found again by the address it ended up at, but no longer by the link you saved it through. A link inside your copy to its own address — the original's address written at its foot — still goes to the web, since it is your way back to the original.
Which pages open this way. Only a page whose address MojoPad can vouch for: one MojoPad fetched from that address itself — Import This Page, File ▸ Import URL…, the bookmarklet with nothing selected, a board's Add web page here…, Explore's Save to wiki, or a PubMed record — or one the Clip to MojoPad extension captured from your own tab. A page whose address only came with it never does: a selection the bookmarklet sends (any web page can send MojoPad a selection, with any address it likes), a web archive you import from a file, a wiki you bring in with Merge Document…, a VoodooPad document, or an assistant's import. Such a page keeps its address, and its ⇱ still opens the site; a link to that address goes to the site. Every page you saved before version 1.99.0 opens this way too, however it came in. A page you send to another wiki with Move to Another Document… or Copy to Another Document…, or into a new one with File ▸ Split Document…, opens this way there exactly when it did in the wiki it came from — so a page taken out of a wiki somebody sent you goes by what that wiki's own file says of it.
To have web links open in your browser as they did before, turn off Settings ▸ Links & Pages ▸ Open your saved copy when a link goes to a web page you saved. A DOI, PubMed or PubMed Central link to a paper you keep still opens your copy, as it always has, and the right-click's Open Your Saved Copy stays.
Choose Import This Page on a link — or File ▸ Import URL… — for a page you already saved, and MojoPad asks before it sends anything to the site: You already have this page, with its name and the day you saved it. Open shows your copy. Save Again fetches the page once more and keeps both copies, for when the page has changed since. Cancel does neither. Nothing is fetched unless you choose Save Again. A board's Add web page here… asks the same, and there the answer that keeps your copy is Add the One I Have, which puts its card on the board; the bookmarklet with nothing selected asks it too.
A report cites a paper by its page on PubMed, and what you want of it is the record: who wrote it, where it appeared, and what it found. Three cases this is for:
How it works. Right-click a link to a paper on PubMed — in a page, or in an Ask answer — and choose Import This Page from PubMed…, or choose File ▸ Import URL… and paste the address. A board's Add web page here… does the same and puts the paper's card on the board, and so does the bookmarklet with nothing selected, on a paper's page in your browser. MojoPad does not load the web page. It asks PubMed's catalog — the one ⌕ Look up this source asks — for the paper's record and its abstract, and sends nothing but the paper's number, with MojoPad's own name and contact address, which PubMed asks every program to give. Since no page is loaded, the question MojoPad asks before it fetches from a site is not asked here: the catalog is asked the way ⌕ Look up this source asks it. The bookmarklet is the one exception: any web page can send MojoPad a clip, so before a clip brings a record in, MojoPad asks that question for the page's own site, as it did when it loaded the page, unless you already allowed that site in this wiki. And a no you gave that question stands: if you said no to MojoPad fetching from PubMed's site in this wiki — pubmed.ncbi.nlm.nih.gov, PubMed Central's pmc.ncbi.nlm.nih.gov, NCBI's www.ncbi.nlm.nih.gov, or the catalog's own eutils.ncbi.nlm.nih.gov — MojoPad sends PubMed nothing and says so, naming the site. That answer lasts until you quit MojoPad; once you open it again, the record can be brought in. The record arrives as a page named by the paper's title, of the kind Source that ⊕ Add sources uses, with its authors, year, journal, volume, issue, pages, DOI, PubMed number and PubMed Central number as properties, each marked as PubMed's answer. Its words are the abstract, with its parts labeled as PubMed labels them — Background, Methods, Results, Conclusions — and the publisher's copyright line; a record that has no abstract says so. The ⇱ at the top opens the paper on PubMed. As with any page you import, Import This Page leaves you where you were and says where the page went, with Open it, and File ▸ Import URL… takes you to it.
A link to an article on PubMed Central works the same way, and its row reads Import the PubMed Record…: what comes in is PubMed's record and abstract, not the article's full text. To keep the whole article, save its PDF or clip it from your browser.
Already in your wiki. When the paper is here already — as its PDF, a source page, or a record you brought in before, found by its PubMed or PubMed Central number or by its address — MojoPad says Already in your wiki, with Open it, and asks PubMed nothing. For a PubMed Central address it first asks PubMed for the paper's PubMed number, and says so when that number finds your copy. A paper you keep under its DOI alone is found by the DOI in the record PubMed sends: the record has been fetched by then, and the note says nothing was imported. Once the record is in, a click on a link to the paper opens it, where a paper opens, and the link is followed by (saved), in a page and in an Ask answer alike.
Add Its Abstract. A page ⊕ Add sources made for a paper has its details and no words. When the paper found is such a page — a formatted text page of yours with nothing at all in it, no words and nothing placed in it such as a picture, a formula, a video or a citation — the note says Its page is empty and offers Add Its Abstract instead of Open it. Choose it, and MojoPad asks PubMed for the record just as above and writes the abstract into that page, labeled as on a record you bring in, keeping the page as it was in Page Versions first. Its name, its details and its kind stay as they were. One thing changes besides its words: as a record you bring in is, the page is marked as having arrived, because its words are PubMed's now, so nothing on it ever runs as a script, a scriptlet you type there later included. A page with anything in it, a page with a password of its own, and a locked page are never written into: they get Open it, and if one of them changes before you press Add Its Abstract, or while PubMed answers, MojoPad says so and writes nothing. A record that has no abstract adds nothing, and the note says so.
When nothing comes. If PubMed does not answer, is busy, or has no record with that number, or you said no to its site in this wiki, MojoPad says which, and nothing is imported. Any other address on PubMed, such as a search, imports as a web page, as before.
Right-click a web link in an Ask answer — in its words, or a web result on the line of sources under it — for Import This Page from its site, Open Your Saved Copy when you have the page, Open in Browser and Copy URL. A model wrote that address, so the menu says which site it is on, and nothing is fetched until you choose Import This Page; then the same question about fetching from that site, and the same refusal of addresses on your own network, stand in front of it as for any page you import. On a link to a paper on PubMed the row reads Import This Page from PubMed…, and it brings in the record as above. A plain click on one opens your copy when you have it, as in a page — and, as in a page, such a link is followed by (saved), or by (PDF) when your copy is a paper you keep, and so is a page named in the answer that is a PDF of yours. A web result on the line of sources wears the same word, after its name; a chip for one of your own pages does not, since it is your page already. The word is drawn, never part of the answer: copy or drag any part of the answer, and it is not in what you paste.
Editing an automatically-created link makes it yours — MojoPad's auto-linking won't rewrite it afterwards. Web destinations may be web, mail, or MojoPad addresses; anything else is refused, so a link can never smuggle in something that runs.
Edit ▸ Copy Link to Page (⌃⇧⌘C) puts a mojopad:// link to
the current page on the clipboard. Paste it into Mail, Things, Calendar, a Terminal
script, anywhere — clicking it opens MojoPad, opens the right document, and lands on
the page. Paste one into another MojoPad document and it auto-links there too:
cross-document links.
Clip the page you're reading straight into MojoPad. Two ways:
What a clip keeps that it used to drop. A page's own head matter comes across now — the title block a document carries at its top, the kind a paper or a report puts above its first paragraph. It used to be cut with the furniture, so a clipped paper arrived starting at its own introduction with the title, authors and date missing.
And what it leaves out on purpose: anything the page is not drawing. The extension captures what is actually on screen, so a collapsed section, an inactive tab, or the rest of an article behind a read more is not in the clip — because it was not in the page when you clicked. Open what you want kept before you clip. Text that is merely faded out as you scroll is kept: it is drawn, just styled to look otherwise.
Pages built out of web components clip properly. Some sites assemble an article from custom elements rather than ordinary markup, and the article's own words live inside them. Those pages used to come through nearly empty. They come through whole now with the browser extension, which reads what the page actually draws. The bookmarklet's whole-page clip has MojoPad fetch the page itself, which cannot see inside those elements, so for a site built this way use the extension.
Every print dialog on your Mac gains a Save PDF to MojoPad entry (in the PDF pop-up menu, bottom-left). Receipts, confirmations, articles — straight into the frontmost document as a viewable PDF page.
Two more mojopad:// links turn any launcher — Raycast, Alfred,
Shortcuts, even the terminal — into a capture tool:
mojopad://today?text=Call%20the%20plumber — appends the text as a new line at the
bottom of today’s daily note (creating it if needed).mojopad://new-page?name=Meeting%20notes&text=First%20line — creates a page
(with optional starting text) and opens it.Try it in Terminal: open "mojopad://today?text=hello". Text arrives as plain
text, exactly as written — remember to URL-encode spaces and punctuation. Both act on the
document window you used last, and need a document open.
The whole scheme, for anyone writing these rather than copying them. Every menu item above puts one of these on your clipboard, so you rarely have to compose one by hand — but a script, a launcher, or a program that files things into your wiki may want to, and here is all of it. Everything in a link is URL-encoded.
| Link | What it does |
|---|---|
mojopad://open?… | Opens a document and lands somewhere in it. The parameters are below. |
mojopad://today?text=… | Adds the text as a new line at the bottom of today's daily note. |
mojopad://new-page?name=…&text=… | Makes a page and opens it. |
mojopad://clip?url=…&title=…&text=… | What the bookmarklet sends. The Web Clipper section above is the door most people want. |
mojopad://manual | Opens this manual. |
The parameters of open. doc says which document, and the rest
say where to land in it. Give as many as you like: naming a page and a folder opens the page
and reveals its folder in the sidebar, which is the pair most worth knowing about.
| Parameter | What it names | Where to get one |
|---|---|---|
doc | The full path to the .mojopad document. If the
document is no longer there, MojoPad looks for it rather than giving up — see below. | Every link MojoPad copies carries it. |
page | A page, by its id — never by its name, so a page you rename keeps every link anyone has made to it. | Edit ▸ Copy Link to Page. |
folder | A folder in the sidebar, by its id. Its ancestors open and it is scrolled into view. | Copy Link to This Folder, right-clicking the folder. |
frag | A place within the page. Four forms, and you never have
to write one: a heading by its words; ^ and a block's id; = and a marked
passage's name, which is what Copy Link to Selected Words writes on a page; and
hl: and a highlight's id, which is what it writes in a PDF. The first two are the
forms a [[Page#…]] link uses inside the wiki. | Copy Link to Selected Words for the last two; type the first two by hand if you are composing a link in a script. |
card | A card on a board, by its id. The board opens with the camera on it. | Copy link to card, right-clicking the card. |
A link whose document has moved still works. A link carries the document's location, so moving or renaming a wiki used to break every link anyone had made to it, silently — clicking one simply did nothing. It now looks for the document instead: a link names a page, a page's id is its filename inside the document, so MojoPad asks the wikis it knows about which one holds that page. That works on links written years ago, and on a wiki that was moved and renamed, where nothing about the old address is true anymore. When it opens a document the link did not name, it says so in the status bar.
It will not guess between two. If you keep a copy of a wiki — a duplicate, a restored backup, the same wiki on two disks — more than one can hold the page, and MojoPad asks you which rather than picking the most recent. If it finds nothing, it offers to let you find the document once; and if the document is simply on a disk that is not connected, it says that instead, because that is not a wiki that has gone anywhere. A wiki MojoPad has never opened at its new location is one it cannot look inside, so the first time you move a wiki, opening it once is what teaches MojoPad where it went.
A link that names something no longer here says so rather than doing nothing: a page or a folder that has been removed, or a marked passage that was edited away, all leave a note in the status bar with the document open in front of you.
One more scheme leaves the app: a citation you copy out of a reader — a page, a page of a
PDF, a marked passage — is written into Markdown and Word exports as a mojoref: link,
and a Word export writes every link to one of your pages, or to a passage you marked, the same way.
It is made by MojoPad and read by MojoPad, so that an exported document can still point back at what
it came from. The one part worth writing yourself is the words of a passage: a link that carries
them marks that passage in the paper when it is clicked — see Link a report to the passages it
quotes, in Transclusion and Panes.
A connected assistant is handed these links ready-made, too: for a page it adds to the wiki you have open, and for each passage you marked that it reads. Its links name the page and not the wiki, since only MojoPad itself can write that part, so a click finds the page in the wiki in front, or in the one wiki of yours that holds it, asking before it opens that one.
A citation says which wiki it came from. This matters most when you keep the same notebook twice — a copy or a restored backup — because both hold the page, and a citation that named only the page could not tell them apart. It now names the document as well, so a quote in a finished manuscript goes back to the wiki it was actually taken from rather than to whichever one you had open. If that wiki is not open, MojoPad asks before opening it; if it cannot be found and its disk is not connected, it says so rather than answering from another copy.
It names the document as a fingerprint, never as a folder. A citation travels — into an export, and from there into a paper you send someone — and a paper that spells out where you keep your work has published something you did not mean to publish. The fingerprint is made with a key that never leaves this Mac, so nobody can check a guess about your folders against it, and only this Mac can match it to a wiki. It does show that two citations came from the same wiki, because they carry the same fingerprint. Citations written before this, citations whose wiki has since moved, and citations followed on another Mac still work the way they always did: MojoPad looks for the page.