Folio changelog

What's new in Folio.

A plain, dated record of the improvements we ship for researchers, students, and institutions.

Release notes

    • The editor's export menu can now draft a journal cover letter and a reviewer-nomination slate in one click, right beside Where to submit? Both are built entirely from the manuscript already in the editor: the significance paragraph is quoted from its own opening rather than summarized for you, and the "why this journal" line states Journal Match's own counts rather than an invented reason. Reviewer names come from the sources the manuscript actually cites, ranked by how often it leans on each. Anything the generator can't know — a note to the editor, the submission confirmations — comes back as a placeholder to fill in, never a guess.
    • A Discovery Agent now learns a little from what you dismiss: after you pass on three or more results from the same venue, similar results from that venue rank lower in later runs — never hidden outright, just pushed down — so repeatedly rejecting one journal's papers stops it from crowding out everything else.
    • Fixed low-contrast text in dark mode across the marketing site, most visibly the primary call-to-action buttons and the cookie banner's Accept button, which were hard to read against their fill in dark mode.

    • Radar topic trackers with more than one word now match the phrase rather than each word on its own, across OpenAlex, arXiv, Europe PMC and DOAJ, and records dated years into the future no longer reach the feed. A tracker like “retrieval practice” sorted by citations now shows the field, not every famous paper that mentions either word.
    • What is new from Radar, and what a Discovery Agent kept, now render as a card inside a Claude or ChatGPT conversation through the MCP connector: each paper with in-library, open-access and retraction badges, Save to Folio, and Not relevant on an agent’s finds to keep a paper out of future runs.
    • An assistant connected through the MCP connector can now set up Radar for you when you ask: follow a topic, an author, a journal or the papers citing one of your sources, create a Discovery Agent for a question and run it on the spot. Namesake authors come back as a choice rather than a guess, plan limits are the same as in Folio, and removing a tracker or an agent stays in Folio. In the connector’s cards, copying a citation or saving a paper now tells the assistant what you did, so the conversation can carry on from it.
    • Three fixes to the connector from testing it against a real account: search and library results now say whether a source has stored text or a stored PDF, so an assistant knows what it can quote or cite before it tries; Radar's live catalog check inside a conversation is now capped so it stays fast; and survey results on a project with a translated instrument no longer list every question twice — responses are attributed to their own instrument version and labeled by language, and a version nobody answered doesn't show up at all.

    • Institutions can now connect an identity provider — Azure AD, Okta, Google Workspace, or any other SCIM 2.0 source — to keep staff and student accounts in sync automatically. A person added, changed, or removed there is reflected in Folio without a manual roster edit, and removing someone revokes their active Folio session immediately.
    • "Write up these results" turns a finished Surveyor study into the start of your paper: it creates a manuscript draft pre-filled with the study's context, its measures, and the number of complete responses, so a Results section starts from the actual data instead of a blank page.

    • A course's public catalog page is now indexable with full course details and one call to action to enroll, and a completion certificate can be added to a LinkedIn profile or issued as an Open Badges credential.