Search university course materials, your flashcards, quizzes, streak and quota. All tools read-only.
Scored across 52 tools
Descriptions work hard to delimit boundaries (e.g. search_library vs list_my_quizzes, get_document vs get_document_content), which raises the ceiling. But several pairs remain genuinely easy to confuse: get_document vs get_document_metadata, get_lecture vs get_lecture_audio, and the cluster of job-status tools (get_quiz_job, get_audio_overview, get_mind_map, get_import_status, get_upload_job) all behave similarly with different parameters.
The vast majority follow a snake_case verb_noun pattern (get_*, list_*, search_*, export_*, calculate_*, find_*, create_*). A few outliers such as billing_status break the verb-first convention, but the overall scheme remains predictable.
52 tools is far past the point where an agent can comfortably hold the surface in mind, and it spans many loosely related domains (billing, referrals, API keys, scopes, support tickets, streaks, study content). Scoping this into separate servers (study content, account/billing, social/community) would serve agents better.
Coverage across documents, lectures, quizzes, flashcards, mind maps, audio overviews, comments, and account admin is broad, with upload/create/list/get flows present. Minor lifecycle gaps exist (e.g. no obvious delete_document, no create_flashcard_set or create_comment visible, billing limited to status/list/upgrade), but agents can work around them.
Get the user's plan, subscription status, renewal or end date and trial state. Pass refresh=true to re-read the subscription from the billing provider first (use it right after a purchase).
Inputs: refresh.
Work out what score is needed on the final exam: current is the average so far (percent), weight is how much the final is worth (percent of the course), target is the course grade wanted (percent).
Inputs: target (required), weight (required), current (required).
Compute a GPA from course grades on a grading scale (letter grades like 'A-' or percentages like 84). scale is a scale id such as standard-4-0 or university-of-toronto; credits default to 1. No account data is read or stored.
Inputs: courses (required), scale.
Export a flashcard set. format=tsv returns Anki-importable text inline; format=pdf returns a one-click download link.
Inputs: setId (required), format (required).
Get a one-click link to download a quiz as a printable PDF with an answer key.
Inputs: quizId (required).
Look up finished documents by a name fragment or a pasted uNotes document link. For topic search across the library use search_library.
Inputs: keyword (required), excludeDocumentId.
Find documents whose file content is identical to a given document (by file hash), for spotting duplicates.
Inputs: documentId (required).
Check an audio overview: pass the jobId from create_audio_overview, or a documentId / lectureSessionId for the latest one. Once COMPLETED it includes the dialogue transcript and a time-limited audioUrl.
Inputs: jobId, documentId, lectureSessionId.
Get the user's spaced-repetition state (due date, stability, difficulty) for every card in one flashcard set they can read.
Inputs: setId (required).
Open one of the user's chat threads with all its messages. includeFileUrl=true also returns a time-limited link to the file attached to the thread, when there is one.
Inputs: threadId (required), includeFileUrl.
Open one comment with its author and the document it is on. Comments on public documents are public.
Inputs: commentId (required).
Open one course (name, code, school, the user's favorite/like flags). withStats=true adds how many documents of each type it holds. Use list_course_documents for its documents.
Inputs: courseId (required), withStats.
Get the metadata of a single uNotes document by ID: title, type, course, school, year, season, professor, language, popularity counters, and its public URL. Use get_document_content to read the actual text.
Inputs: documentId (required).
Get the full text content of a specific document, lecture session, or flashcard set. Use this when you need to read the actual content of a material to answer questions about it.
Inputs: documentId, flashcardSetId, lectureSessionId.
Check whether the user can download a document's original file today (their plan's daily limit) and get the one-click link to the document page, where the Download button is. The download itself counts against the daily limit when the user presses it.
Inputs: documentId (required).
Get the school and course a document belongs to.
Inputs: documentId (required).
Get the user's spaced-repetition review queue: flashcards due now across all sets (oldest due first), plus how many are due and how many were learned today. limit=0 returns counts only.
Inputs: limit.
Get one flashcard set the user owns, with all its cards (prompt and answer).
Inputs: setId (required).
Check a Quizlet import started with import_quizlet.
Inputs: jobId (required).
Open one of the user's lectures: title, status, notes, transcript and recording details.
Inputs: lectureId (required).
Say whether a lecture has a recording (format, length) and get a one-click link to the lecture page, where it plays.
Inputs: lectureId (required).
Check a mind map: pass the jobId from create_mind_map, or a documentId / lectureSessionId for the latest one. Once COMPLETED it includes the concept tree.
Inputs: jobId, documentId, lectureSessionId.
Get the user's profile: name, username, email, whether a password is set, email-subscription state, notification categories and plan.
Inputs: no named parameters.
Load an existing quiz of a course (or one named quiz of that course) so the user can take it again, with its questions and answer options. Spends no AI quota.
Inputs: courseId (required), quizId.
Check a quiz generation started with generate_quiz: pass the documentId (latest job for that document) or the jobId.
Inputs: jobId, documentId.
Get the authenticated user's plan tier and current usage against each quota (AI chat, quiz generation, flashcard generation, uploads, downloads, transcription minutes, audio overviews, mind maps). A null limit means unlimited.
Inputs: feature.
Show the user's Refer & Earn state: their personal invite link, clicks, signups, sales and earnings (pending, cleared, paid), plus their invite points. includeReferrals=true also lists the people they referred.
Inputs: includeReferrals.
Open one school (name, abbreviation, course-code prefixes). Use list_schools_courses to browse its courses.
Inputs: schoolId (required).
Get the authenticated user's daily study streak: current streak length, longest streak, whether they already studied today (UTC), and milestone badges (7/30/100/365 days).
Inputs: no named parameters.
Check the progress of an upload started with upload_document or upload_document_from_url. Pass the jobId, or leave it out to list the user's uploads that are still running.
Inputs: jobId.
List the user's active API keys (name, prefix, permissions, last used). The key value is never shown again after creation.
Inputs: no named parameters.
List the user's AI chat threads (newest first, with messages). source='brain' (default) is the study-assistant chat that ask_brain writes to; source='copilot' lists the in-app copilot conversations. Filter by scope or by the document, lecture or flashcard set a chat is about.
Inputs: scope, source, documentId, flashcardSetId, +1 more.
List the AI apps connected to this uNotes account (with what each may do and when it last called) and the sign-in providers linked to it.
Inputs: no named parameters.
List the documents of one course (use list_schools_courses to find the course id), optionally filtered by name, season or year, or get the course's document counts by type.
Inputs: courseId (required), name, skip, year, +3 more.
List the user's favorite courses, documents or schools. Optionally filter by name and page with skip and limit.
Inputs: entityType (required), skip, limit, query.
Search the public flashcard library (not just the user's own sets). query matches title and subject.
Inputs: skip, limit, query.
Show which permission groups and scopes this connection currently holds.
Inputs: no named parameters.
List the user's recorded lectures, newest edit first, optionally filtered by title.
Inputs: skip, limit, query.
List the comments the user has written, newest first, each with the document it is on. Page with skip and limit.
Inputs: skip, limit.
List the courses or the schools the user created, optionally filtered by name.
Inputs: type (required), name, skip, limit.
List the documents the user uploaded. scope=all (default) filters by name, season, year or type; processing shows uploads still being processed; stats and counts give totals by document type; quiz_sources lists the documents a quiz can be generated from.
Inputs: name, skip, type, year, +3 more.
List the authenticated user's own flashcard sets on uNotes, newest first. Use it for any question about the user's flashcards; optionally filter by title, subject, or course code. The document library search does not cover them.
Inputs: skip, limit, query.
List the authenticated user's own quizzes on uNotes (quizzes for the courses where the user has uploaded documents), newest first. Use it for any question about the user's quizzes, such as "what quizzes do I have for CSI3131?": put the course code or quiz title in query. This is the tool for the user's quizzes; the document library search does not cover them.
Inputs: skip, limit, query.
Page the user's notification history, newest first.
Inputs: skip, limit.
List the paid plans (Premium, Pro) with their billing periods and prices. Use the id with upgrade_plan.
Inputs: no named parameters.
Search the public quiz library by course, school or text (all published quizzes, not just the user's).
Inputs: skip, limit, query.
Search for schools and courses on uNotes. Use this when the student needs to find a school or course for uploading a document or creating flashcards.
Inputs: type, limit, query, schoolId.
The ways the user can sign in to this uNotes account (Google, Apple, email and password), whether each one can be disconnected and why not, and the ones not connected yet. The last way in is never removable. Connecting a provider is a browser sign-in with that provider: give the user connectUrl. A password is set through request_password_reset.
Inputs: no named parameters.
List the support messages this user has sent, newest first.
Inputs: limit.
Ask for more permissions. Returns a one-click approval link for the user; after they approve, the new permissions apply on the next call. Connections made with an API key have fixed permissions: create a new key instead.
Inputs: tiers (required).
Search the public uNotes document library by title, course, or type keywords. Call it with the query alone whenever no scope id is already known — schoolId and courseId are optional narrowing hints, never prerequisites, and must never be requested from the user. Course codes and school names the user typed belong in the query text. It searches documents only: it is not for the user's own quizzes or flashcards, which have their own listing tools.
Inputs: query (required), limit, courseId, schoolId.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Search, read, cite, create, and safely update a user's private KeepFlash knowledge library.
Search and read the articles, videos, podcasts, PDFs and notes you saved in Recall. Read-only.
Search and read your own notes: insights, drafts, book and podcast notes. Read-only, OAuth.
Read-only search of Youth4work public jobs, assessed talent, exam mock tests and Indian colleges.
Add one secure layer between your agents and this server.
