note2it-browse
ChatGPTWalk the notebook → section → page hierarchy. With no arguments, lists every notebook the user can access (their own plus shared read-only ones) with section and page counts. With notebook_id, returns that notebook's full section tree and the pages at each level, in display order. Add section_id to show just that section's subtree. Use this to discover where things live or to pick a destination; use note2it-search when you're looking for content. Ids returned here (notebook:/section:/page:) are what the other tools accept.
note2it-create-from-template
ChatGPTInstantiates one of the user's templates into a new, filled-in page. Pass values as a map of token key → value (keys exactly as returned by note2it-list-templates): scalar tokens take a string; {{list:…}} tokens take an ARRAY of item strings (a newline-separated string also works and is split for you), and render using the bullet/number/checkbox style set on the template token itself. Hashtags inside a value (e.g. a {{categories}} token filled with "#mystery #thriller") become real, searchable tags. Missing tokens fill as empty — the response lists any you left blank so you can patch them later. {{img:…}} tokens take the media key or URL that note2it-upload-image returns — upload the image first with that tool, then pass its reference as the token's value; a key or URL you don't own is rejected. Destination defaults to the template's own notebook/section; override with the same destination parameters as note2it-create-page. The template itself is never modified.
note2it-create-page
ChatGPTCreate a new page from markdown. Destination: pass notebook_id (+ optional section_id), or notebook / section_path by name (e.g. notebook: "Work", section_path: "Projects/Active") — names must match existing containers unless create_missing: true, which creates the notebook and/or section path (notebook creation is subject to the user's plan limit; the error tells you when it's reached). Markdown supports headings, lists, checkboxes, tables, fenced code, quotes, and images by URL. Write [[Page Title]] to link to another page (a graph edge) and #tag for searchable tags; write {{token}} literally only when authoring a template (with is_template: true). Returns the new page reference. To start from one of the user's templates, use note2it-create-from-template instead.
note2it-fetch
ChatGPTRetrieve one page by reference (page:<id> or bare 24-hex id) or by exact title. The default format: "markdown" returns the page as markdown with [[wiki-links]], #tags, and {{template tokens}} preserved as literal text — safe to read and quote. The response header shows the page reference, its notebook, tags, outgoing [[wiki-links]], and the last-edited timestamp. format: "outline" returns the block skeleton (one line per block: id, type, first characters) — useful to understand structure. If a title matches several pages, the candidate list is returned instead of guessing — pick one and re-fetch by its page:<id>. Use note2it-search when you don't know the exact title.
note2it-links
ChatGPTExplore the wiki-link graph around a page. direction: "in" lists backlinks — every page whose text contains [[This Page's Title]]. "out" lists the pages this page links to, flagging links whose target page doesn't exist yet (broken or pending links). "both" (default) returns the full neighborhood. depth: 2 expands one more hop around each neighbor — use it to map a topic cluster. Wiki-links resolve by title, so this also accepts a bare title that has no page yet (answering "what links to [[X]]?" before X exists). Use note2it-fetch to read any page this returns.
note2it-list-templates
ChatGPTLists every page the user has flagged as a reusable template, across all of their notebooks, with each template's fill-in tokens already parsed: token key, human label, and input kind (text, multiline, date, number, or image for {{img:…}} placeholders). Use it to tell the user what templates they have and which values a template needs. To see a template's full layout, note2it-fetch it like any page; to create a page from one, use note2it-create-from-template.
note2it-move-page
ChatGPTMove a page to another notebook and/or section (by id, or by name with the same resolution rules as note2it-create-page, including create_missing), or change its position among its siblings (position: 0 = top). Moving never alters content, links, or tags; wiki-links are title-based so the graph is unaffected, and the page's tasks follow it to the new notebook. You can only move pages you own, into notebooks you own. Omitted destination fields keep their current value; section_id: null moves the page to the notebook root.
note2it-page-history
ChatGPTThe undo story for connector edits. Without restore_revision_id, lists the page's saved revisions (newest first): every MCP write snapshots the page beforehand, so each entry is the page as it was BEFORE that edit. With restore_revision_id, restores that version's title and content — the current state is snapshotted first, so a restore is itself undoable. Restoring does not move the page or change its trash state.
note2it-patch-page
ChatGPTApply targeted edits without touching the rest of the page — the safe default for editing. First call note2it-fetch with format: "outline" to get block ids and updatedAt. Operations (applied atomically in order): append / prepend (markdown at the end/start of the page); insert_after / insert_before / replace_block / delete_block (by block id from the outline); append_to_section / replace_section (anchored on a heading's exact text — the section runs to the next heading of the same or higher level; occurrence disambiguates duplicate headings; replace_section keeps the heading itself); replace_text (exact-string replacement inside one styled text run, preserving formatting — if your target spans formatting boundaries the error will say so; use replace_block then). Markdown in operations converts [[links]] and #tags like create does. if_updated_at is required; on a 409 conflict, re-fetch and re-apply. Unknown block ids and headings return the available options so you can self-correct. Every patch is snapshotted server-side first — edits are undoable.
note2it-search
ChatGPTFull-text search across every page the user can read (their own notebooks plus notebooks shared with them), matching titles and body text — including text inside tables, image captions, map pins, and calendar events. Returns up to limit results ranked by relevance, each with the page reference (page:<id>), notebook, tags, a short excerpt, and last-edited date. Narrow with filters: notebook_id/section_id scope the hierarchy, tag restricts to pages carrying that #tag, is_template finds (or excludes) templates, updated_after finds recent work. An empty query with at least one filter acts as a filtered listing (e.g. all pages tagged #project). Results are summaries only — use note2it-fetch with the page reference to read a page in full. If you already know the exact title or id, prefer note2it-fetch directly.
note2it-tags
ChatGPTThe user's tag index: every #tag in use, with page counts and up to three most-recently edited pages per tag. Use it to answer "what topics do I have?", to pick an established tag before suggesting one (avoid near-duplicates like #proj vs #project), or as the entry point before note2it-search with a tag filter for the full page list.
note2it-trash-page
ChatGPTSoft-deletes a page: it disappears from notebooks, search, the graph, and templates, but stays in trash and is fully restorable with restore: true. This tool never permanently deletes anything — permanent deletion stays an owner-only action in the note2it app. If other pages link to the trashed page via [[wiki-links]], the response lists them so you can clean up or warn the user. Trashing a template removes it from the template picker.
note2it-update-page
ChatGPTWhole-page operations: rename, replace the entire body with new markdown, or toggle the template flag. Prefer note2it-patch-page for any partial change — this tool replaces the whole body and is only right when rewriting a page wholesale. When renaming, update_backlinks (default true) rewrites every [[Old Title]] reference across the user's other pages so the graph never breaks; the response lists which pages were updated. Pass if_updated_at (from your last fetch) to fail safely if the page changed since you read it. The previous version is snapshotted server-side before every write. This tool cannot move pages (note2it-move-page) or delete them (note2it-trash-page).
note2it-upload-image
ChatGPTDownloads an image from a public https URL (png/jpeg/webp/gif/avif) and stores it in the user's note2it media library, counted against their plan's file-count and storage quota. The image is fetched and validated server-side (not by the connector), so the URL must be publicly reachable over https. Returns the stored media key and its note2it-hosted URL — use that URL to fill an image:… template token via note2it-create-from-template, or paste it into markdown as an image block.
note2it-browse
Claudenote2it-create-from-template
Claudenote2it-create-page
Claudenote2it-fetch
Claudenote2it-links
Claudenote2it-list-templates
Claudenote2it-move-page
Claudenote2it-page-history
Claudenote2it-patch-page
Claudenote2it-search
Claudenote2it-tags
Claudenote2it-trash-page
Claudenote2it-update-page
Claudenote2it-upload-image
Claude