Revise

See what changed in a document since YOU last looked at it — e.g. edits the human (or another collaborator) made in between. Returns added, removed, and changed blocks (with block ids), all in the accepted/projected state. Great for catching up before continuing work ("how does it look now?"). The first call on a document just establishes a baseline. Each call advances your baseline to the current state.

get_recent_changes

Read the host user's current semantic selection from a live Revise editor tab. Returns selected text, caret/range positions, portable target segments, active formatting, comment IDs, and pending suggestion IDs, with no viewport geometry. The document must currently be open in the user's browser.

get_selection

Create a Revise document from a file at a public http(s) URL (.md, .markdown, .txt, .html, .htm, .docx, .rtf, .odt; PDFs/images not yet supported). The server fetches the URL — file bytes are never passed inline. For a LOCAL file, use upload_document instead (it streams the file to the server). Returns the new document id and URL.

import_document

Insert a footnote: a numbered superscript reference placed immediately after the anchor text inside a paragraph, plus the footnote body (shown at the bottom of the page, or as an endnote in pageless documents). On substantial documents this lands as a pending tracked change by default, like edit_document. Footnote bodies appear in read_document as [^n]: definitions with their own block ids — edit or remove them with the normal editing tools (removing a footnote body also removes its reference, and vice versa).

insert_footnote

Share a document with a specific person by email, as a read, write, or suggest collaborator (suggest = can only propose tracked changes and comments). Existing Revise users are added directly; others receive an email invite. (Agents cannot make documents publicly accessible — that's a human-only action in the app.)

invite_collaborator

List the user's Revise documents (most recently edited first). Returns id, title, sharing state, and timestamps.

list_documents

List the documents the user currently has open in a live Revise editor (a browser tab) right now — useful for joining them in what they're actively working on. Returns an empty list when nothing is open. Reflects the human's own browser sessions only, not documents you (the agent) are reading through these tools.

list_open_documents

Show a document to the user as a live, editable view in the conversation. Use it when the user wants to see or work in an existing document. Don't call it after create_document: creating a document already shows this view. The view updates as you edit, so don't open the document again after each change.

open_document

Read a Revise document's content. Format "markdown" (default) is the simple dialect — best for plain prose; set with_block_ids to true to interleave <!-- block:xxxxxx --> anchors usable with edit_document. Format "html" is the full-fidelity dialect: compact HTML-like markup with a block id on every element plus everything markdown cannot express — rich marks, <latex> math, code block languages, merged table cells, page layout, and each comment thread inline as a <comment-thread transcript="..."> wrapper around its anchored text. Prefer html when a document uses rich features or has comments; edit_document accepts the same dialect in replacements. "text" is plain text. (For a styled, self-contained HTML file, use export_document instead.) view controls how pending tracked-change suggestions read: "final" (default, as if accepted) or "original" (as if rejected). Long documents are paginated: when a read exceeds the character budget it is cut at a block boundary and the response carries truncated: true, next_start_block, and a ready-to-run example call — repeat with start_block to continue. For a targeted read of a large document, prefer get_document_outline + search_within_document over paging through everything.

read_document

Used by the document view itself to renew its connection. Not for the model.

get_document_widget_session

Change a document's title.

rename_document

Replace the ENTIRE content of a document with new markdown. Destructive: existing content is removed (it remains recoverable via the document's revision history). Prefer edit_document for targeted changes.

replace_document_content

Full-text search across the user's Revise documents. Returns matching document ids with short match excerpts — use search_within_document to find WHERE in a document the matches are. Search indexing is asynchronous, so a document created or renamed moments ago may take a few seconds to appear; use list_documents for immediate title lookup.

search_documents

Find where a query matches inside one document. Returns the matching blocks as markdown with <!-- block:xxxxxx --> anchors (usable with edit_document), each with a few neighboring blocks for context. Case-insensitive.

search_within_document

Patch a document's page or section layout: page size, orientation, margins, 1-3 columns, page numbers, line/paragraph spacing, pageless mode, or the APA-7 preset. ONLY the fields you pass change; everything else keeps its current or inherited value. Pass section_index to change one section; omit it to apply the passed fields to the document defaults and every section. Length values are strings with explicit units (e.g. "1in", "2.54cm", "72px").

set_page_layout

Style existing content with forgiving CSS-lite selectors and values. Use this for bulk formatting instead of rewriting text. Block selectors include *, #blockId, p, h1-h6, li, lists, tables, and descendants. Inline selectors such as b/strong, i/em, u, s/strike/del, code, sup, sub, mark, and a target only matching formatted text ranges. Examples: set the entire document to Georgia with selectors="*" and attrs=[{"name":"fontFamily","value":"Georgia"}]; bold headings with selectors=["h1","h2","h3"] and attrs=[{"name":"fontWeight","value":"bold"}]. Formatting lands as pending tracked changes by default on substantial documents.

style_blocks

Create a Revise document from a LOCAL file (.md, .markdown, .txt, .html, .htm, .docx, .rtf, .xml, .odt; PDFs/images not yet supported) without putting its bytes in model context. This tool requires the MCP host to read the local filesystem and run a shell command with network access (Codex, Claude Code, and ChatGPT Work mode have this; browser-only clients and sandboxes without network do not have shell access that can reach the server). A file the user attached to a chat in the Claude or ChatGPT app (a path like /root/.claude/uploads/... or /mnt/user-data/...) is in a sandbox that can't reach this server: don't call upload_document for it. Instead: If your client can show interactive app views (UI widgets) in the chat: call create_document with a short title (such as the file's name) and no content, which opens it in the Revise document view, and tell the user to click Upload on its empty page (or drag the file onto it). The file fills that document, which keeps its id, and you are told when it's in. If your client cannot show app views: tell the user to open https://revise.io/docs/upload, upload the file there (docx, md, txt, html), and paste the resulting revise.io/docs/... link back into this chat. If the curl cannot run, follow the if_curl_unavailable guidance in the response. It returns a one-time upload_url and ready-to-run curl command that streams the file and returns the new document's id and url. For a file already at a public http(s) URL, use import_document instead.

upload_document