Patsnap Patent & Literature Search

Search 200M+ patents and 216M+ scientific papers in natural language with Patsnap.

От сообществаРаботаетБез входаГлобальныйБесплатноТолько чтение

Что умеет

  • Patsnap Fetch: # Patent or Paper Markdown Fetch Tool ## Tool Description This is a data retrieval tool designed for AI Agents to fetch patent and paper data. It supports batch retrieval of Markdown-fo
  • Patsnap Search: # Patent / Paper Search Search one source per request (`source`: `patent` default, or `paper`). Returns a ranked page of matching documents with metadata and snippets. ## Use for Findi

Какие данные видит

Нужен ли аккаунт

Не нужен: сервер работает без входа

Search 200M+ patents and 216M+ scientific papers in natural language with Patsnap. Supports patent search, literature search, fusion search, semantic and keyword queries, and precise filters for assignee, inventor, IPC, legal status, jurisdiction, dates, and citations.

Как подключить

Как подключить Patsnap Patent & Literature Search к Claude

Адрес сервера, который нужно вставить в настройках Claude:

https://patsnap-patent-literature-search--openpatsnap.run.tools

  1. Откройте Claude (сайт claude.ai или приложение для компьютера).
  2. Перейдите в Настройки → Коннекторы (в английском интерфейсе: Settings → Connectors).
  3. Нажмите Добавить пользовательский коннектор (Add custom connector).
  4. Вставьте адрес сервера, скопированный ниже, в поле URL удалённого MCP-сервера и нажмите Добавить.
  5. В чате нажмите + → Коннекторы и включите новый коннектор.

Пользовательские коннекторы доступны на тарифах Free, Pro, Max, Team и Enterprise. На Free можно добавить только один. В Team и Enterprise коннектор сначала добавляет владелец организации в Настройки организации → Коннекторы.

Авторизация

После нажатия Добавить откроется окно входа в сервис. Войдите под своим аккаунтом и подтвердите доступ. Пароль Claude не видит.

Как подключить Patsnap Patent & Literature Search к ChatGPT

Адрес сервера, который нужно вставить в настройках ChatGPT:

https://patsnap-patent-literature-search--openpatsnap.run.tools

  1. Откройте ChatGPT в браузере (chatgpt.com). Нужен тариф Plus, Pro, Business, Enterprise или Edu.
  2. Один раз включите режим разработчика: Настройки → Приложения → Дополнительные настройки → Режим разработчика (Settings → Apps → Advanced settings → Developer mode).
  3. Откройте Настройки → Коннекторы и нажмите Создать (Create).
  4. Заполните форму: Название (любое), Описание (одна фраза, что делает сервис), URL MCP-сервера (скопируйте ниже).
  5. В поле Аутентификация выберите OAuth, если сервис требует вход, иначе Нет. Нажмите Создать.
  6. В новом чате откройте + → Приложения/Коннекторы и включите коннектор.

Название раздела у OpenAI менялось (Connectors → Apps/Plugins). Если пункт называется иначе, ищите «режим разработчика» в настройках. В рабочих пространствах Business/Enterprise администратор должен разрешить пользовательские коннекторы.

Авторизация

При первом использовании ChatGPT откроет окно входа в сервис. Войдите и подтвердите доступ.

Как подключить Patsnap Patent & Literature Search к Cursor

Адрес сервера, который нужно вставить в настройках Cursor:

https://patsnap-patent-literature-search--openpatsnap.run.tools

Быстрый способ: нажмите кнопку Открыть в Cursor ниже, Cursor сам предложит добавить сервер.

Вручную:

  1. В Cursor откройте Settings → Cursor Settings → MCP и нажмите Add new global MCP server.
  2. Вставьте JSON, скопированный ниже, в открывшийся файл ~/.cursor/mcp.json (для одного проекта: .cursor/mcp.json в корне проекта).
  3. Сохраните файл. Сервер появится в списке MCP; при необходимости пройдите авторизацию по кнопке рядом с ним.

Авторизация

Если сервис требует вход, рядом с сервером в списке MCP появится кнопка авторизации.

Список инструментов сервера (2)

Технические названия из tools/list. Нужны только разработчикам.

patsnap_fetch# Patent or Paper Markdown Fetch Tool ## Tool Description This is a data retrieval tool designed for AI Agents to fetch patent and paper data. It supports batch retrieval of Markdown-formatted content based on a list of keys (URLs containing patentId or paperId, or patent publication numbers), enabling AI to perform content analysis, reasoning, and Q&A. ## Core Features - **Batch Query**: Supports querying one or multiple patents or papers at once (up to 100) - **Auto Detection**: Automatically identifies whether the input is a patent URL (containing patentId), a paper URL (containing paperId), or a patent publication number (pn) - **Format Conversion**: Converts structured data into readable Markdown format - **Image Handling**: Optionally includes image links (for patents) - **Content Compression**: Optionally compresses returned content to reduce transfer size - **Module Selection**: Optionally specify which data modules to retrieve (basic, citations, legal, family) - **Structured Data**: Optionally returns the raw structured JSON data alongside Markdown, grouped by module ## Input Parameters ### keys (required) - **Type**: Array of strings - **Description**: List of keys (URLs containing patentId or paperId, or patent publication numbers) - **Patent URL format**: e.g. `https://eureka.zhihuiya.com/view/#/fullText'figures/?patentId=07c38692-9ccf-4ead-8f0f-484f7aa929b7` - **Paper URL format**: e.g. `https://eureka.zhihuiya.com/literature/#/abst?paperId=c53a8bcb-b7a3-42c9-8a6f-a26d6b5dec9a` - **Patent publication number (pn) format**: e.g. `US20240368139A1` - **Count limit**: 1–100 (10–20 recommended for best performance) ### key_type (optional) - **Type**: String - **Default**: `"url"` - **Description**: Key type — either a URL (containing patentId or paperId) or a patent publication number (pn) - **Usage**: Set to `"url"` when the input keys are URLs; set to `"pn"` when the input keys are patent publication numbers - **Allowed values**: `url`, `pn` ### compress (optional) - **Type**: Boolean - **Default**: `false` - **Description**: Whether to compress the returned Markdown content (using gzip + base64 encoding) - **Usage**: Recommended when the returned content is large ### include_images (optional) - **Type**: Boolean - **Default**: `false` - **Description**: Whether to include image links (only applicable to patents) - **Note**: When enabled, image URLs will be automatically signed ### module (optional) - **Type**: Array of strings - **Default**: `["basic"]` - **Description**: List of data modules to retrieve (only applicable to patents) - **Allowed values**: - `basic` — Core patent info (biblio, title, abstract, claims, description, etc.) - `citation` — Citation and cited-by records - `legal` — Legal status, assignment, license records - `family` — Patent family members (original, INPADOC, extended) - **Note**: `citation`, `legal`, and `family` are only supported for patents, not papers - **Examples**: `["basic"]`, `["basic", "citation"]`, `["basic", "citation", "legal", "family"]` ### include_structured (optional) - **Type**: Boolean - **Default**: `false` - **Description**: Whether to return the raw structured JSON data (the `structured` field) alongside the Markdown content - **Usage**: Enable when you need the original source data for further processing ## Output ### Successful Response Returns a JSON object with the following fields: - `total`: Total number of keys submitted - `success_count`: Number of successfully converted items - `failed_count`: Number of failed items - `results`: List of results, each containing: - `key`: The original input key (URL or patent publication number) - `key_type`: The type of the original input key (`url` or `pn`) - `markdown`: The converted Markdown content - `images`: List of image info (only present for patents when `include_images` is enabled) - `success`: Whether the conversion was successful - `message`: Processing message - `source`: Data source (`patent` or `paper`) - `modules`: The list of modules queried for this item - `structured`: Raw structured JSON data grouped by module (`basic`, `citation`, `legal`, `family`); unqueried modules are `null`. Only present when `include_structured` is `true` ## Usage Examples ### Example 1: Query a Patent ```json { "keys": ["https://eureka.zhihuiya.com/view/#/fullText'figures/?patentId=07c38692-9ccf-4ead-8f0f-484f7aa929b7"], "compress": false, "include_images": true, "modules": ["basic", "citation", "legal", "family"] } ``` ```json { "keys": ["CN115885075B"], "key_type": "pn", "compress": false, "include_images": true, "modules": ["basic"] } ``` ### Patent Result: ```json { "total": 1, "success_count": 1, "failed_count": 0, "results": [ { "key": "https://eureka.zhihuiya.com/view/#/fullText'figures/?patentId=07c38692-9ccf-4ead-8f0f-484f7aa929b7", "key_type": "url", "markdown": "# 专利详细信息\n\n...", "images": [ { "image_id": "HDA0003320126060000011", "image_url": "https://example.com/image.png" } ], "success": true, "compress": false, "message": "获取成功", "source": "patent", "modules": ["basic", "citation", "legal", "family"] } ] } ``` ### Example 2: Query a Paper ```json { "keys": ["https://eureka.zhihuiya.com/literature/#/abst?paperId=c53a8bcb-b7a3-42c9-8a6f-a26d6b5dec9a"], "compress": false, "include_images": false, "modules": ["basic"] } ``` ### Paper Result: ```json { "total": 1, "success_count": 1, "failed_count": 0, "results": [ { "key": "https://eureka.zhihuiya.com/literature/#/abst?paperId=c53a8bcb-b7a3-42c9-8a6f-a26d6b5dec9a", "key_type": "url", "markdown": "# 论文详细信息\n\n...", "images": [], "success": true, "compress": false, "message": "获取成功", "source": "paper", "modules": ["basic"] } ] } ```
patsnap_search# Patent / Paper Search Search one source per request (`source`: `patent` default, or `paper`). Returns a ranked page of matching documents with metadata and snippets. ## Use for Finding or listing patents/papers by topic, technical problem, company/applicant, author/inventor, identifier, classification, jurisdiction, date range, legal status, or citation threshold — for review, comparison, or ranking. ## Do NOT use for - Counts, trends, distributions, top-N, or citation statistics — this tool returns documents, not aggregates. - Full document content (full text, claims, descriptions, images, family, citation lists, legal detail) or raw source records. - Chit-chat or general knowledge unrelated to patents/papers. ## Invariants (MUST obey) 1. **Strategy ⇄ params binding is strict.** - `"semantic"` in `search_strategy` → MUST provide `semantic_query`; if `"semantic"` is NOT in `search_strategy` → MUST NOT provide `semantic_query`. - `"keyword"` in `search_strategy` → MUST provide `keywords`; if `"keyword"` is NOT in `search_strategy` → MUST NOT provide `keywords`. - `"filter"` in `search_strategy` → MUST provide `filters` with at least one field set; if `"filter"` is NOT in `search_strategy` → MUST NOT provide `filters`. 2. **`semantic_query` is for natural-language conceptual search only.** - Use it for a technical question, problem, mechanism, or solution description. - Do not use it as a keyword list, company lookup, product lookup, identifier lookup, date slice, jurisdiction slice, IPC slice, or legal-status slice. 3. **`keywords` are atomic topic terms for BM25 text matching.** - Each item should be one technical term, product/standard name, abbreviation, method, material, or research concept. - Do not pass sentences, comma-joined strings, or generic filler such as "technology", "method", "system", "development". - Do not put applicant/company/assignee, inventor, author, or organization names in `keywords`; those belong in `filters`. 4. **`filters` are structured constraints, not text keywords.** - Use `filters` for company/applicant/assignee, inventor/author, organization, identifier, jurisdiction, IPC, legal status, date range, and citation threshold constraints. - Only set fields the user explicitly states. Do not infer dates, jurisdiction, IPC, assignees, inventors, authors, organizations, legal status, or citation thresholds. - `filters` must be a JSON object, never a stringified/escaped JSON string. 5. **Pagination limits are fixed.** - `limit` 1–100, `offset` ≥ 0, `offset + limit` ≤ 1000. ## Filter fields - `source="patent"`: pn, apno, assignees, inventors, jurisdiction, ipc, legal_status, date_from, date_to, date_type, cited_min - `source="paper"`: authors, org_names, date_from, date_to, cited_min ## Examples ```json {"source":"paper","search_strategy":["semantic"],"semantic_query":"pharmaceutical technical solutions for treating glaucoma","limit":20} ``` ```json {"source":"patent","search_strategy":["keyword","filter"],"keywords":["5G","antenna","MIMO"],"filters":{"assignees":["Huawei"],"jurisdiction":["CN","US"]},"limit":50} ``` ```json {"source":"paper","search_strategy":["filter"],"filters":{"authors":["Zhang San"],"date_from":20230101,"date_to":20241231},"limit":30} ```