Lexpad is a vocabulary notebook with spaced repetition, in 81 languages. Connect it to ChatGPT or Claude and ask the assistant to save a word from your chat, show the words due today or quiz you. Each answer counts as practice and moves the word through its Leitner boxes. Sign-in is OAuth 2.1 with PKCE on a Lexpad page. 20 tools; the ones that overwrite or delete are marked, so the client asks first. Connecting is free. On Lexpad's Free plan the assistant writes 10 cards a day; on Pro it is unlimited, up to 1,000 a day.
Guide and example prompts: https://lexpad.app/features/assistants
Scored across 20 tools
The four read tools (get_due_words, get_recent_words, list_words, search_words) substantially overlap: list_words with filter=due or sort=recent reproduces the other two, and the descriptions spend most of their length explaining the differences. The word/meaning mutation tools (add_word, update_word, add_meaning, update_meaning, remove_meaning, reorder_meanings) are cleanly separated, so overlap is real but contained.
Nearly every tool follows a verb_noun snake_case pattern (add_word, update_meaning, set_starred, record_answer, list_notebooks). A few outliers break the pattern: lexpad_help, notebook_stats (noun-only) and whoami (no separator). The convention is still readable and predictable overall.
Twenty tools sits at the heavy end of the range, and the count is inflated by the redundant retrieval trio plus an ancillary help/feedback/whoami group that is unrelated to the core vocabulary workflow. Each tool is defensible on its own, but the surface could be consolidated to roughly 12-15 without losing capability.
Word and meaning lifecycles are fully covered (create, read, update, delete, reorder), alongside practice recording, scheduling state (starred/retired), stats, search and help. The main gap is notebook management: list_notebooks exists but there is no create/rename/delete notebook or any export/import operation, leaving some lifecycle dead ends.
Adds one meaning, with examples, in the notebook's meaning language, to a word already in the notebook. It only adds; nothing is replaced or removed.
Inputs: meaning (required), word_id (required). Returns: plan, word, added
Saves one word or phrase the user asked to save to their notebook, with meanings in the notebook's meaning language and examples in its target language. It takes a whole card: several meanings, each with a gloss and examples, part of speech, level, pronunciation, collocations, expressions, related words, a topic and notes; everything but the headword is optional. If the notebook already has the word, it moves to the top and only new meanings are added. A word saved with a meaning uses one of the account's daily AI cards for assistants; when none are left, it is saved without one and the result says when they come back.
Inputs: headword (required), pos, level, topic, +12 more. Returns: plan, word, existing, meaningsAdded
Deletes one word from the user's notebook, with its meanings and their practice progress, on every device. It cannot be undone here. Needs the separate words:delete permission.
Inputs: word_id (required). Returns: id, deleted, headword, notebookId
The words waiting for review today: due today or overdue, by the date in the user's own timezone, in one notebook or all. Paused words and words not yet practised are left out. Newest added first. It takes no filter, search or sort, so it is the ready-made review list. Each word comes with a short form of its meanings, its box and its due date. Up to 100 words a page (default 50). The first page gives total, the number of matching words, counted up to 10000 (totalIsAtLeast marks a larger count). While more remain, a page ends with next_cursor; sent back as cursor with the same arguments, it returns the next page.
Inputs: limit, cursor, notebook_id. Returns: total, words, nextCursor, notebookId, +1 more
Every word in one notebook, or all, in the order it was added, newest first, whatever its progress: new, learning, mastered and paused alike. It takes no filter, search or date range; it answers what the user saved lately. Each word comes with a short form of its meanings, its status and the day it was added. Up to 100 words a page (default 50). The first page gives total, the number of matching words, counted up to 10000 (totalIsAtLeast marks a larger count). While more remain, a page ends with next_cursor; sent back as cursor with the same arguments, it returns the next page.
Inputs: limit, cursor, notebook_id. Returns: total, words, nextCursor, notebookId, +1 more
One word's full card: every meaning with examples, collocations, notes, and the user's progress (box, due date, right and wrong answers, their own note), which schedule its notebook uses (boxes or smart) and, in a smart notebook, the chance in percent that the user remembers it today.
Inputs: word_id (required). Returns: id, pos, level, topic, +14 more
Answers questions about Lexpad itself from Lexpad's own public pages: what it can do, how to do something in the app, platforms and downloads, privacy, plans, prices and allowances, and what changed lately. A search returns up to 4 short passages, each with the lexpad.app page it comes from. Topic plans adds the AI card and Lex allowances, the trial length and, outside ChatGPT, the prices in force right now; topic whats_new returns the newest changes with their dates. found is false when the pages say nothing about it. The content follows the website, so it covers features added after this tool was listed.
Inputs: limit, query, topic. Returns: found, plans, source, changes, +4 more
The user's notebooks with their ids, languages, word counts and which one is the default.
Returns: notebooks
Browses the user's words by any mix of: a filter (all; due, waiting for review today; new or not_started, not yet practised; struggling, keeps going back to box 1; starred; mastered; retired, paused in the app; corrections; hard_spelling), a CEFR level, a topic, a date range (added_after, added_before), a text that narrows the list, and a sort (recent, oldest or alphabetical). notebook_id all covers every notebook. It lists matches in the chosen order, not ranked by how well they match. Each word comes with a short form of its meanings and its progress. Up to 100 words a page (default 50). The first page gives total, the number of matching words, counted up to 10000 (totalIsAtLeast marks a larger count). While more remain, a page ends with next_cursor; sent back as cursor with the same arguments, it returns the next page.
Inputs: sort, level, limit, query, +6 more. Returns: total, words, nextCursor, notebookId, +1 more
Per notebook: how many words, due today, new, mastered, struggling, starred and hard to spell; and the user's streak. dueToday is every word due; today is how many of them the app's practice asks today under the notebook's reviews a day (reviewsPerDay, 0 = no limit), the number the app's Today screen shows.
Inputs: notebook_id. Returns: streak, notebooks, longestStreak, moreNotebooks, +1 more
Records whether the user answered a practice question about a word correctly. It counts like a review in the app: it moves the word's box, the day's numbers and the streak, and the result gives the word's next review date. A right answer to a word that is not due yet may be saved as practice, which keeps its next review (scheduleChanged false). Each answer carries its own client_id, so the same answer sent twice counts once.
Inputs: correct (required), word_id (required), client_id (required). Returns: box, leech, status, streak, +3 more
Removes one meaning, with its examples, from a word, chosen by its position in the word's card. It cannot be undone.
Inputs: word_id (required), meaning_index (required). Returns: id, pos, level, topic, +14 more
Puts a word's meanings in a new order. The first one is the main meaning, the one practice asks about. order lists every current index once, e.g. [2, 0, 1] makes the third meaning the main one.
Inputs: order (required), word_id (required). Returns: id, pos, level, topic, +14 more
The best matches for a query in one notebook, in either language, best first. It returns at most 50 and does not page or filter.
Inputs: query (required), limit, notebook_id. Returns: words, notebookId
Sends a note from the user to the Lexpad team: a problem, an idea, a question or praise about Lexpad, with an optional category. It is stored with the feedback people send from the app's Settings and read by the team. Only the text, the category and which assistant sent it are kept. Each account may send a few notes a day.
Inputs: message (required), category. Returns: id, category, createdAt
Pauses a word (the app calls it Paused): it stops coming up in practice and keeps its progress. Or brings a paused word back into practice.
Inputs: retired (required), word_id (required). Returns: id, pos, level, topic, +14 more
Stars or unstars a word, as the star on its card does. Starred words have their own practice deck.
Inputs: starred (required), word_id (required). Returns: id, pos, level, topic, +14 more
Changes one meaning of a word, chosen by its position in the word's card: its text, its gloss, or its examples (a list given replaces them). Fields not given stay as they are. The previous text is replaced.
Inputs: word_id (required), meaning_index (required), gloss, meaning, +1 more. Returns: id, pos, level, topic, +14 more
Changes a word's headword, part of speech, level, topic, pronunciation, hard-to-spell mark and spelling note, usage note, corrected form, collocations, expressions, related words or the user's private note. Only the fields given change; meanings and practice progress are not touched. A headword the notebook already has is refused. The previous values are replaced.
Inputs: word_id (required), pos, note, level, +10 more. Returns: id, pos, level, topic, +14 more
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Free: turn your AI chats into spaced-repetition vocabulary. 13 tools, reads and writes.
Connects an AI assistant to your Lingard account (lingard.app), an app for learning languages with books, news and flashcards. Create topic word collections and add complete cards: translation, part of speech, pronunciation, synonyms, example sentences with translations, and images. Fill in missing fields on existing cards, rewrite example translations, and pause or resume collections, books and news in your daily study sessions. Every change can be previewed before it is saved, and saved words appear in the Lingard apps on iOS, Android and the web. Requires a Lingard account; sign-in uses OAuth.
Save vocabulary from any AI chat; see each word in every language you know; review it with FSRS.
Create and review spaced-repetition flashcards in your free Klappa decks from any AI assistant.
Add one secure layer between your agents and this server.
