GraphChinese: Chinese dictionary and grammar

Chinese dictionary (124k entries), HSK levels, text segmentation and 195 grammar points.

我該用這個嗎

品質與安全性

A
說明品質
100%
結構描述完整度
97%
命名品質
96%
汙染風險
100%
權限相符程度
100%
協定合規性
100%

根據工具定義與協定合規性的自動化分析。

上下文成本

~1,840Token(工具定義)
~683 B典型回應大小
中等的注意力影響(128k 上下文的 1.44%)

這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高,可用於其他工作的注意力就越少。

安裝

一鍵安裝

將以下內容加入你的 `claude_desktop_config.json` 檔案:

{
  "mcpServers": {
    "chinese-dictionary-grammar": {
      "url": "https://graphchinese.com/mcp"
    }
  }
}

遠端端點

https://graphchinese.com/mcpstreamable-http

它能做什麼

工具清單

工具(10)

🟢 唯讀🟡 寫入🔴 刪除⚪ 未知
🟢search_dictionary(query)

Search a Mandarin Chinese dictionary by Chinese characters, pinyin or an English word. Returns up to 10 matching words, each with its pinyin, a short meaning, its HSK level when it has one, and a link to its dictionary page. Use this when the user wants the Chinese for an English word, or does not know the exact headword. Do not use it for whole sentences (use segment_text) or grammar questions.

輸入結構描述

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "Chinese characters, pinyin (with or without tones) or English.",
      "minLength": 1,
      "maxLength": 64
    }
  },
  "required": [
    "query"
  ],
  "additionalProperties": false
}
🟢lookup_word(word)

Look up one Chinese word or character, written in simplified or traditional characters. Returns its pinyin, every meaning, its HSK level, other readings of the same characters, each character it is made of, and the GraphChinese lessons that teach it. Use this when the user asks what a specific Chinese word means or how it is pronounced. Do not use it for Japanese or other languages.

輸入結構描述

{
  "type": "object",
  "properties": {
    "word": {
      "type": "string",
      "description": "The word in Chinese characters, for example 便宜.",
      "minLength": 1,
      "maxLength": 32
    }
  },
  "required": [
    "word"
  ],
  "additionalProperties": false
}
🟢get_related_characters(character)

Show how one Chinese character connects to others. Returns the components it is built from, characters that contain it, characters that look similar and are easy to confuse with it, the other characters that share its sound component, and common words that start with or contain it. Use this when the user wants to learn a character through its parts, tell look-alike characters apart, or see a character family. Do not use it for words of two or more characters.

輸入結構描述

{
  "type": "object",
  "properties": {
    "character": {
      "type": "string",
      "description": "Exactly one Chinese character, for example 请.",
      "minLength": 1,
      "maxLength": 4
    }
  },
  "required": [
    "character"
  ],
  "additionalProperties": false
}
🟢segment_text(text)

Split a Chinese sentence or short passage into words. Returns each word in order with its dictionary pinyin, a short meaning and its HSK level, plus a count of words per HSK level. The pinyin is each word's dictionary reading, not a reading chosen for this sentence. Use this to gloss a sentence word by word or to judge how hard a text is. Do not use it to translate a whole passage.

輸入結構描述

{
  "type": "object",
  "properties": {
    "text": {
      "type": "string",
      "description": "Chinese text, up to 1000 characters.",
      "minLength": 1,
      "maxLength": 1000
    }
  },
  "required": [
    "text"
  ],
  "additionalProperties": false
}
🟢search_grammar(query, level)

Search 195 Mandarin grammar points by keyword, by HSK level, or both. A keyword can be English (comparison, passive), a Chinese marker (把, 了) or a term name. Returns matching points with a one-paragraph summary and the slug that get_grammar_point takes. Use this to find which grammar point covers a pattern. Do not use it for vocabulary questions.

輸入結構描述

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "One or more keywords. Every keyword must match.",
      "minLength": 1,
      "maxLength": 64
    },
    "level": {
      "type": "integer",
      "minimum": 1,
      "maximum": 6,
      "description": "Only return points taught at this HSK level."
    }
  },
  "required": [],
  "additionalProperties": false
}
🟢get_grammar_point(slug)

Get one Mandarin grammar point by its slug. Returns the written explanation, worked example sentences with pinyin and English, a word-by-word note for each word in the pattern (what it does, structure, three examples, a common mistake), the grammar points that should be learned first, and the points this one leads to. Use this after search_grammar, when the user wants a pattern explained or wants to know what to study before it. Do not guess a slug: get it from search_grammar or get_grammar_order.

輸入結構描述

{
  "type": "object",
  "properties": {
    "slug": {
      "type": "string",
      "pattern": "^[a-z0-9][a-z0-9-]*$",
      "maxLength": 80,
      "description": "A slug returned by search_grammar."
    }
  },
  "required": [
    "slug"
  ],
  "additionalProperties": false
}
🟢get_study_order(level)

Get the opening vocabulary lessons of an HSK level, in the order GraphChinese teaches them, each a themed group of words with pinyin and meaning. This is a learning order, not an alphabetical word list. Use this when the user asks what words to learn first, or what to start with at HSK 1 to 5. Do not use it to look up one word (use lookup_word).

輸入結構描述

{
  "type": "object",
  "properties": {
    "level": {
      "type": "integer",
      "minimum": 1,
      "maximum": 5,
      "description": "HSK level, 1 to 5."
    }
  },
  "required": [
    "level"
  ],
  "additionalProperties": false
}
🟢get_common_characters(count, offset)

List the most frequent Chinese characters, most common first, each with its pinyin, meaning, HSK level and the cumulative share of running text it covers, plus how many characters are needed to cover 50, 80, 90 and 95 percent. Counted over GraphChinese's 145 graded readings. Use this when the user asks which characters matter most or how many they need. Do not use it for one character's meaning (use lookup_word).

輸入結構描述

{
  "type": "object",
  "properties": {
    "count": {
      "type": "integer",
      "minimum": 1,
      "maximum": 50,
      "description": "How many characters to return. Default 20."
    },
    "offset": {
      "type": "integer",
      "minimum": 0,
      "maximum": 99,
      "description": "Skip this many of the most common. Default 0."
    }
  },
  "required": [],
  "additionalProperties": false
}
🟢get_topic_words(topic)

Get the core Chinese words for a topic such as food, family, travel or business, with pinyin, meaning and HSK level. Call it with no topic to list the 30 topics and their slugs. Use this when the user wants words for a situation or subject. Do not use it to translate a specific word.

輸入結構描述

{
  "type": "object",
  "properties": {
    "topic": {
      "type": "string",
      "pattern": "^[a-z][a-z-]*$",
      "maxLength": 40,
      "description": "A topic slug such as food. Omit to list all topics."
    }
  },
  "required": [],
  "additionalProperties": false
}
🟢get_grammar_order(level, count)

List the grammar points of an HSK level in the order GraphChinese teaches them, which respects what each point needs first. Each has a summary and the slug that get_grammar_point takes. Use this when the user asks which sentence patterns to learn first, or what grammar to study at their level. Do not use it to explain one pattern (use get_grammar_point).

輸入結構描述

{
  "type": "object",
  "properties": {
    "level": {
      "type": "integer",
      "minimum": 1,
      "maximum": 6,
      "description": "HSK level, 1 to 6."
    },
    "count": {
      "type": "integer",
      "minimum": 1,
      "maximum": 20,
      "description": "How many points to return. Default 10."
    }
  },
  "required": [
    "level"
  ],
  "additionalProperties": false
}

社群

為此伺服器評分

證據

近期觀測

已驗證未記錄版本10 個工具