Hand the dictionary to your LLM.

Look up a single Chinese character (traditional). Returns Taiwan frequency rank, pinyin (multiple readings when applicable), bopomofo, English definition, TOCFL level, radical, stroke count, and common words built on this character (example_words, derived from word_chars).

List traditional Chinese characters filtered by rank range, radical, or stroke count, ordered by Taiwan (COCT) frequency. Use for queries like 'top 20 characters', 'all characters with the 木 radical', or '7-stroke characters'.

Look up a single traditional Chinese word — a multi-character vocabulary item such as 電腦 or 謝謝. Returns Taiwan frequency rank, pinyin, bopomofo, English definition, TOCFL level and part of speech, plus the full record for each component character, so the word can be explained from its parts. Use get_character for a single glyph.

List traditional Chinese words ordered by Taiwan (COCT) frequency — the vocabulary side of the database. Filter by frequency rank, or by a character the word must contain: `contains: 學` returns the words built on 學, which is how a learner turns one known character into usable vocabulary. Lists only words carrying a frequency rank; for a graded level's complete vocabulary use list_tocfl_terms.

List the Kangxi radicals in use — the semantic components characters are built from, and the traditional way of indexing a dictionary. Each entry carries the radical glyph, its Kangxi ordinal (1-214), a one-word English gloss (木 → tree), and how many characters in the inventory use it. Radicals with no characters are omitted. Pass a glyph to list_radical_characters, or to the `radical` filter on list_characters.

Takes no arguments.

List the characters built on one Kangxi radical, most useful first — Taiwan frequency, then stroke count. The response also carries the radical's ordinal and English gloss. Use this to answer 'what else uses 水?' or to teach a radical as a group. It differs from list_characters' `radical` filter in that it includes characters the frequency corpus never attests, and orders by stroke count within equal frequency, which is how a paper dictionary arranges them.

List the 7 TOCFL (華語八千詞) proficiency levels — Taiwan's official graded vocabulary scale, traditional script. Returns each level's slug, display name, and entry count (characters + words). Pass the returned slug to list_tocfl_terms. Note the display names are NOT the ordinals: 1=Novice 1, 2=Novice 2, 3=Level 1, 4=Level 2, 5=Level 3, 6=Level 4, 7=Level 5.

Takes no arguments.

List the vocabulary at one TOCFL level — characters and/or multi-character words from the 華語八千詞 list (traditional script), ordered by Taiwan frequency (unranked entries last). Each entry carries pinyin, bopomofo, English definition, part of speech, and frequency rank. Call list_tocfl_levels first to get the level slug. Use this for queries like 'TOCFL Level 1 words' or 'beginner Taiwan vocabulary'.

Find characters and words by meaning or pronunciation, when you don't have the glyph: English ("computer"), pinyin with or without tone marks or numbers ("diannao", "diàn nǎo", "dian4nao3"), or the traditional characters themselves. Results are ordered by match quality — exact reading first, then a definition that leads with the searched meaning, then Taiwan frequency. Reach for this before asserting how something is said or written in Taiwan; use get_character / get_word instead when you already have the exact term. Lexical search, not semantic: it matches readings and definition wording, so a near-synonym the gloss doesn't use won't be found.