Vocab Voyage
Measured conclusion: Deep-tested. L2 outcome: tools_ok.
20 MCP tools + 17 widgets for SAT/ISEE/SSAT/GRE/GMAT/LSAT prep. Flashcards, quizzes & games. Hosted.
| Author | io.github.jaideepdhanoa |
|---|---|
| Access | remote |
| Publisher key required | No |
| Measured state | Deep-tested |
| L1 liveness | alive |
| L1 latency | 410 ms |
| Last measured | 2026-10-24 |
Add to your agent
Copy this into your MCP client config:
{
"mcpServers": {
"Vocab Voyage": {
"url": "https://gponcrussdahcdyrlhcr.supabase.co/functions/v1/mcp-server?ref=mcp_registry",
"transport": "streamable-http"
}
}
}
Measured tools (31)
get_word_of_the_dayread-onlyUse this when the user asks for today's word, a daily vocabulary nudge, or a single-word warmup. Returns today's deterministic Word of the Day (definition, part of speech, example, synonyms/antonyms), optionally scoped to a test family (isee, ssat, sat, psat, gre, gmat, lsat, general). Do not use for arbitrary lookups — call get_definition instead.
Parameters:
test_family— string — Optional test family: isee, ssat, sat, psat, gre, gmat, lsat, general
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_definitionread-onlyUse this when the user asks what a specific word means, requests its definition, part of speech, synonyms/antonyms, or an example sentence. Returns curated dictionary data from the Vocab Voyage corpus. Do not use for sentence-level meaning disambiguation (call explain_word_in_context) or for daily word prompts (call get_word_of_the_day).
Parameters:
word(required) — string — The word to define
Hints: read-only yes · destructive no · idempotent yes · open-world no
generate_quizread-onlyUse this when the user wants to practice, be quizzed, or test their knowledge across multiple words at once. Generates a 1–10 question multiple-choice quiz for a test family (isee, ssat, sat, psat, gre, gmat, lsat, general). Renders the interactive Vocab Voyage quiz widget on supporting hosts; per-answer taps persist mastery for signed-in users. Do not use for definition lookups — call get_definition instead. Do not use for spaced-repetition flashcards — call get_flashcards instead.
Parameters:
test_family(required) — stringcount— number — Number of questions (1–10), default 5level— string — Optional difficulty hint
Hints: read-only yes · destructive no · idempotent no · open-world no
get_course_word_listread-onlyGet a sample of vocabulary words from a specific Vocab Voyage course. Use list_courses to discover slugs.
Parameters:
course_slug(required) — stringlimit— number — 1–50, default 20
Hints: read-only yes · destructive no · idempotent yes · open-world no
list_coursesread-onlyLists all 13 Vocab Voyage courses with their slugs and descriptions.
Hints: read-only yes · destructive no · idempotent yes · open-world no
explain_word_in_contextread-onlyExplain what a word means inside a specific sentence — useful when a word has multiple meanings.
Parameters:
word(required) — stringsentence(required) — string — The sentence containing the word
Hints: read-only yes · destructive no · idempotent yes · open-world no
study_plan_previewread-onlyUse this when the user asks for a study plan, a multi-day prep schedule, or how to prepare for a test by date. Returns a 7-day plan (5 words/day) for a given test family. Renders the interactive Vocab Voyage study-plan widget on supporting hosts; tapping 'Start Day N' launches a flashcard session seeded with that day's words. Do not use for a single quiz session — call generate_quiz instead. Do not use for one-off lookups — call get_definition instead.
Parameters:
test_family(required) — stringtarget_date— string — Optional ISO date (YYYY-MM-DD)
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_flashcardsread-onlyUse this when the user asks for flashcards, wants to drill words individually, or wants a tap-to-flip review session. Returns 1–12 cards for a test family. Renders the interactive Vocab Voyage flashcards widget on supporting hosts; per-card 'I knew it / I didn't' buttons persist mastery for signed-in users. Do not use for multiple-choice testing (call generate_quiz) or for a single word lookup (call get_definition).
Parameters:
count— number — Number of cards (1–12), default 5test_family— string — isee, ssat, sat, psat, gre, gmat, lsat, general
Hints: read-only yes · destructive no · idempotent no · open-world no
get_my_progressread-onlyUse this when the signed-in user asks about their own streak, XP, words mastered, recent activity, or 'how am I doing'. Auth-only personal dashboard. Renders the interactive Vocab Voyage progress widget on supporting hosts; falls back to markdown elsewhere. Anonymous callers receive a sign-in prompt. Do not use for global stats or other users' progress.
Hints: read-only yes · destructive no · idempotent no · open-world no
play_gameUse this when the user wants to play a vocabulary game, asks for something fun, or wants to learn through play. Launches one of 11 mini-games inside the host chat. Renders the matching ui://vocab-voyage/game/{slug} widget on supporting hosts; falls back to a deep link elsewhere. Per-question answers persist via record_word_result; round completion fires record_session_complete + award_game_xp so MCP play counts toward streaks, XP, and mastery for signed-in users. Supported slugs: word_match, spelling_bee, speed_round, synonym_showdown, word_scramble, fill_in_blank, context_clues, word_guess, picture_match, crossword, word_search. Do not use for a serious test-prep quiz — call generate_quiz instead.
Parameters:
slug(required) — string — Game slug: word_match | spelling_bee | speed_round | synonym_showdown | word_scramble | fill_in_blank | context_clues | word_guess | picture_match | crossword | word_searchcount— number — Words in the round (4–12, default 8)test_family— string — Optional: isee, ssat, sat, psat, gre, gmat, lsat, general
Hints: read-only no · destructive no · idempotent no · open-world no
record_word_resultPersist a single word answer (correct/incorrect) to the user's mastery progress. Mirrors the web app's word-mastery scaling so MCP study counts toward leaderboards and streaks. Requires sign-in.
Parameters:
is_correct(required) — booleanword— string — The word answered (preferred for human input).card_id— string — Card UUID (preferred when known from a prior tool result).question_type— string — e.g. multiple_choice, fill_in_blank, flashcard.quiz_attempt_id— string — Optional quiz_attempt UUID to record a per-question row.selected_answer— stringtime_taken_seconds— number
Hints: read-only no · destructive no · idempotent no · open-world no
record_session_completeRecord a completed study session: writes study_sessions, awards study-time XP (+1/min, capped 30/day), and updates the daily streak. Use after a play_game / quiz / flashcard session ends. Requires sign-in.
Parameters:
cards_studied(required) — numbertime_spent_seconds(required) — numberdeck_id— string — Optional deck UUID; omit for ad-hoc MCP sessions.total_count— numbersession_type— string — e.g. mcp_word_match, mcp_flashcard, mcp_quiz.correct_count— numbersession_title— string
Hints: read-only no · destructive no · idempotent no · open-world no
award_game_xpAward score-based XP from a game/activity (separate from study-time XP). Cascades to the leaderboard via DB trigger. Requires sign-in.
Parameters:
xp(required) — number — XP to award (>= 0).reason— string — Optional human label for analytics.
Hints: read-only no · destructive no · idempotent no · open-world no
mark_word_knownManually mark a word as mastered for the signed-in user (same as the flashcard 'I knew this' override). Requires sign-in.
Parameters:
word— stringcard_id— string
Hints: read-only no · destructive no · idempotent yes · open-world no
mark_word_difficultManually mark a word as still-learning for the signed-in user (resets mastery toward learning band). Requires sign-in.
Parameters:
word— stringcard_id— string
Hints: read-only no · destructive no · idempotent yes · open-world no
update_adaptive_levelRun the adaptive-mastery promotion logic for the signed-in user (delegates to the web app's update-adaptive-mastery function). Requires sign-in.
Parameters:
words_studied(required) — numbercorrect_count(required) — numbertotal_count(required) — numbercourse_id— string
Hints: read-only no · destructive no · idempotent no · open-world no
file_support_ticketFile a real human-followup support ticket on behalf of the signed-in user. Use this when the user reports a bug, account lockout, complaint about a tutor, or anything Sparkle/the agent cannot resolve from data. The ticket is emailed to the support team and a confirmation is sent to the user with a 1-business-day SLA. Vocab Voyage is completely free, so there is nothing to bill — never raise pricing, plans or upgrades. Categories: billing (legacy, use account instead), bug, account, complaint, feedback, other. Requires sign-in.
Parameters:
category(required) — string — Issue category. Use 'bug' for crashes/data loss, 'account' for lockouts/access or any account/access question, 'complaint' for tutor/quality issues, 'feedback' for feature requests. 'billing' is legacy and unused — the product is free.summary(required) — string — One-line description of the issue (what the user needs).conversation_snippet— string — Optional: last few turns of the conversation for context.
Hints: read-only no · destructive no · idempotent no · open-world yes
get_recent_mistakesread-onlyUse this when the signed-in user asks about words they've gotten wrong, missed words, words to review, or wants to revisit recent mistakes. Returns up to 25 words from the last N days (default 7) with miss-rate and last-seen timestamp, plus a link to the in-app Recent Mistakes page. SUMMARISE — never dump every row; tell the user the count, name 2–3 sample words, and recommend the page URL. Requires sign-in.
Parameters:
days— number — Lookback window in days (1–90, default 7)limit— number — Max words to return (1–50, default 10)
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_session_detailread-onlyUse this when the signed-in user asks 'what did I miss in [that session]', 'which words tripped me up', or 'what was my accuracy on session X'. Pass a session_id (study_sessions.id or adaptive_sessions.id, usually obtained from get_recent_session_results / a picker chip). Returns title, accuracy %, wrong_words[] (max 10), and a per-card timeline (truncated to first 20 events). Cite at least one wrong word and the accuracy in your reply.
Parameters:
session_id(required) — string — study_sessions.id or adaptive_sessions.id (UUID)include_timeline— boolean — Include per-card timeline (default true). Truncated to 20 events.
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_pending_invitesread-onlyUse this when the signed-in user asks about pending parent invites, share codes, or whether their parent invite has been accepted yet. Returns each pending invite with hours_until_expiry. RULE: if any invite has hours_until_expiry < 24 (and not expired), proactively offer to resend it via the resend-parent-invite flow. If expired, offer to send a fresh invite. Requires sign-in.
Hints: read-only yes · destructive no · idempotent yes · open-world no
nudge_childParent-only. Sends a 'check-in' push notification (and email fallback) to a linked child. Use when the parent says things like 'remind my kid to study', 'nudge my child', 'tell Sam to do their words today'. The server enforces a 24h cooldown per child — if rate-limited the response includes retry_after_hours. NEVER spoof a different parent — the calling user must already be linked to the child. Requires sign-in.
Parameters:
child_user_id(required) — string — user_id of the linked child to nudgereason— string — Optional short reason (≤200 chars), e.g. 'streak at risk'message— string — Optional personal message (≤280 chars) shown to the child
Hints: read-only no · destructive no · idempotent no · open-world yes
resend_pending_inviteResend a pending parent invite by id. Use after get_pending_invites surfaces an invite expiring in <24h, or when the user explicitly asks to resend. Re-emails the existing invite_token; no new code is generated. 60s per-invite cooldown. Caller must own the invite. Requires sign-in.
Parameters:
invite_id(required) — string — The id field returned by get_pending_invites.
Hints: read-only no · destructive no · idempotent no · open-world yes
get_class_standingread-onlyUse this when a signed-in student asks how they're doing in their tutor class, who's ahead, who their rival is, or who they should challenge. Auth-only. Returns weekly XP rank inside the user's tutor class plus a winnable rival suggestion (similar weekly XP). NEVER name the class leader unless the user is rank #1 — the response uses '(top student)' as a deliberate placeholder. Renders the interactive class-standing widget on supporting hosts; falls back to markdown elsewhere. Anonymous callers receive a sign-in prompt.
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_sparkle_guidanceread-onlyReturns Vocab Voyage's lifecycle-aware guidance: the user's current phase (e.g. student.at_risk), a friendly greeting, 2–3 recommended tool calls, and an optional CTA. Renders the session-debrief widget on supporting hosts. Anonymous callers get visitor.* phase suggestions.
Parameters:
persona— string — Optional persona override: student | parent | tutor | explorer.
Hints: read-only yes · destructive no · idempotent no · open-world no
set_personaread-onlyBias subsequent Sparkle guidance toward a persona (student | parent | tutor | explorer). Session-scoped: the host should pass the chosen persona back to get_sparkle_guidance.
Parameters:
persona(required) — string — student | parent | tutor | explorer
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_recommended_next_actionread-onlyOne-line 'do this next' hint derived from the user's current lifecycle phase. Useful when the agent wants a quick recommendation without rendering a full guidance card.
Parameters:
persona— string — Optional persona override.
Hints: read-only yes · destructive no · idempotent no · open-world no
list_starter_promptsread-onlyLists Vocab Voyage's MCP starter prompts (also exposed via the standard MCP prompts/list endpoint). Useful for hosts that don't yet support prompts/list.
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_session_trendsread-onlyAuth-only. Personal study trends over a window (default 14 days, max 90): session count, total minutes, accuracy trend (up/down/flat), and top-missed words. Use after a user asks 'how am I trending / am I improving / which words keep tripping me up'.
Parameters:
days— number — Window size in days (default 14, max 90).
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_class_session_trendsread-onlyAuth-only. Tutor-only. Aggregate class-level trends across the tutor's classes (default 14 days, max 30). Pass `class_id` to scope to one class; omit it to get a worst-first rollup across up to 25 classes plus 1–3 struggling students.
Parameters:
days— number — Window size in days (default 14, max 30).class_id— string — Optional class id to scope to one class.
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_child_session_detailread-onlyAuth-only. Parent-only. Detailed breakdown for a single child's study/quiz session — accuracy, missed words, duration. Defaults to the most recent session for the parent's first linked child if no `child_user_id` / `session_id` is supplied. Ownership-gated: returns an error for unlinked children.
Parameters:
session_id— string — Optional. Defaults to the child's most recent session.child_user_id— string — Optional. Defaults to first linked child.
Hints: read-only yes · destructive no · idempotent yes · open-world no
get_study_plan_recommendationread-onlyAuth-only. Returns a personalized N-day study plan (default 7, range 3–7) chosen from one of four focus modes (weak-topic-drill / streak-recovery / new-words / review-mastery) based on the user's recent trends. Inline only the first 3 days; full plan persists when the user clicks the Vocab Voyage start link.
Parameters:
horizon_days— number — Plan length in days (3–7, default 7).
Hints: read-only yes · destructive no · idempotent no · open-world no
Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.