SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillslinkfox-zhihuiya-abstract-data-translated

linkfox-zhihuiya-abstract-data-translated

Writing & research

View repository

Updated 10-15 09:05

About this skill

Zhihuiya Patent Abstract (Translated)

This skill guides you on how to retrieve translated patent titles and abstracts from the Zhihuiya (PatSnap) patent database, supporting Chinese, English, and Japanese translations.

Core Concepts

Zhihuiya (PatSnap) is a leading patent intelligence platform. This tool queries its database to return translated titles and abstracts for one or more patents. You can look up patents by patent ID or publication (announcement) number, and receive translations in Chinese, English, or Japanese.

Patent identification: Each patent can be identified by either a patentId (internal Zhihuiya identifier) or a patentNumber (public publication/announcement number such as US20200012345A1 or CN112345678A). If both are provided, the patent ID takes priority.

Family patent fallback: When the original patent has no abstract available, you can optionally substitute the abstract from a related family patent by enabling the replacement option.

Parameter Guide

Parameter Type Required Description
patentId string At least one of patentId or patentNumber Zhihuiya internal patent ID. Separate multiple IDs with commas. Max length 60,000 characters.
patentNumber string At least one of patentId or patentNumber Publication (announcement) number. Separate multiple numbers with commas. Max length 60,000 characters.
replaceByRelated integer No Whether to substitute a family patent abstract when the original is unavailable. 1 = yes, 0 = no. Default 0.
lang string No Target translation language. en = English (default), cn = Chinese, jp = Japanese.

Key Rules

  1. At least one identifier is required: You must provide either patentId or patentNumber (or both). If both are supplied, patentId takes priority.
  2. Batch queries: Multiple patents can be queried at once by separating values with commas.
  3. Default language is English: When the user does not specify a language, use en.
  4. Family fallback: Set replaceByRelated to 1 only when the user explicitly wants a substitute abstract from a family patent if the original is missing.

Response Fields

Field Description
total Number of patent records returned
data Array of patent objects (see below)
data[].patentId Zhihuiya internal patent ID
data[].pn Publication (announcement) number
data[].title Translated patent title
data[].abstractText Translated patent abstract
data[].pnRelated Publication number of the substitute family patent (only present when family replacement was used)
costToken Tokens consumed by this request

Usage Examples

1. Translate a single patent abstract to English by publication number

Look up patent number US20200012345A1 and give me the English abstract.

Parameters: patentNumber = "US20200012345A1", lang = "en"

2. Translate multiple patents to Chinese

Get the Chinese translation of abstracts for patents CN112345678A and US20200067890A1.

Parameters: patentNumber = "CN112345678A,US20200067890A1", lang = "cn"

3. Look up by patent ID with family fallback

Get the Japanese abstract for patent ID 12345678. If the abstract is unavailable, use a family patent instead.

Parameters: patentId = "12345678", lang = "jp", replaceByRelated = 1

4. Batch query by patent IDs

Translate the titles and abstracts for these patent IDs: 111111, 222222, 333333.

Parameters: patentId = "111111,222222,333333", lang = "en"

Display Rules

  1. Present data clearly: Show results in a well-structured table with patent number, title, and abstract.
  2. Indicate language: Mention the translation language in the output header so users know which language the results are in.
  3. Family patent notice: If pnRelated is present in any result, explicitly inform the user that the abstract was sourced from a family patent and show the substitute publication number.
  4. Long abstracts: For very long abstracts, display the full text without truncation so users can review the complete content.
  5. Error handling: Report the actual error. An empty translation result means the requested translation was not returned; it does not establish that the patent is absent or that all same-language requests are unsupported. If the user needs the original abstract, use /zhihuiya/simpleBibliography and verify the returned patent and text language. Keep the requested target language; do not silently switch it or substitute a family patent.
  6. No subjective commentary: Present the translated text as-is without adding interpretation or legal analysis of the patent content.

How to Call

  • API Endpoint：POST /zhihuiya/abstractDataTranslated(for full parameters/responses/error codes, see references/api.md）
  • Python Script：python scripts/zhihuiya_abstract_translated.py '<JSON 参数>' [--inline]
  • Cost Constraints: This tool consumes computing power; by default, the same parameter combination is called only once per session, and the script has a 24h local cache. On failure/empty results, do not automatically try different keywords, page through, or keep probing by changing the postal code; if continued retrieval is needed, first explain to the user that additional consumption will be incurred.

Output Strategy (Default Script Behavior)：

  • Alwayswrite the full response to <cwd>/linkfox/<YYYY-MM-DD>/<session>/data/linkfox-zhihuiya-abstract-data-translated-<timestamp>.json（<cwd> is the working directory when the script is executed, which in Claude Code is the current project directory;<session> is taken from the environment variable SESSION_ID, automatically aggregated according to the user's task;writing to /tmp is prohibited, and if the current directory is not writable, report an error)
  • Response body ≤ 8 KB: after writing to disk, print the full JSON to stdout
  • Response body > 8 KB: after writing to disk, stdout outputs only a summary (top-level fields, common counts such as total/costToken, length of the largest list field + first 3 samples)
  • Add --inline to force full printing to stdout (also written to disk)

Data Reading Recommendations: first look at the summary to determine whether it is sufficient; when specific fields are needed, prioritize using jqorConvertFrom-Json to extract on demand from the saved json file, avoiding bringing the entire JSON into context.

Resolving Authentication and Computing Power Issues

When the following abnormal situations occur, use references/onboarding.md to guide resolution:

Abnormal Situations

  • API Key not configured: the environment variable is not configured LINKFOX_AGENT_API_KEY, and LINKFOXAGENT_API_KEY。
  • Response 401 or 402 status code
  • Response indicates insufficient computing power or balance: the message contains "insufficient computing power balance/insufficient billing/insufficient balance/quota exceeded/insufficient balance/package expired/recharge required/please recharge", or content with a similar meaning.

Important Limitations

  • Identifier required: At least one of patent ID or publication number must be provided; the tool cannot perform keyword-based searches.
  • Translation languages: Only Chinese (cn), English (en), and Japanese (jp) are supported.
  • No full-text retrieval: This tool returns only titles and abstracts, not full patent claims or descriptions.
  • Family replacement is optional: The substitute family patent abstract is only provided when explicitly requested via replaceByRelated = 1.

User Expression & Scenario Quick Reference

Applicable -- Patent abstract and title translation queries:

User Says Scenario
"Translate this patent abstract" Single patent translation
"What does patent XX say / what is it about" Abstract lookup
"Get the Chinese/Japanese version of this patent" Specific language translation
"Look up the abstract for patent number XX" Publication number lookup
"Translate these patents in batch" Batch translation
"The abstract is missing, try a family patent" Family patent fallback

Not applicable -- Needs beyond abstract translation:

  • Full patent text, claims, or description retrieval
  • Patent search by keyword, classification, or applicant
  • Patent legal status, citation analysis, or landscape reports
  • Patent valuation or infringement analysis

Computing Power Consumption Rules

Billed according to dynamic rules: computing power consumed = 81 × number of returned data entries. Each entry is 1 patent abstract translation result

Important: the service of this skill is dynamically calculated by a multiplier and may consume a large amount of computing power at once; the user must be reminded, and the user decides whether to continue.

Feedback:

Auto-detect and report feedback via the Feedback API when any of the following apply:

  1. The functionality or purpose described in this skill does not match actual behavior
  2. The skill's results do not match the user's intent
  3. The user expresses dissatisfaction or praise about this skill
  4. Anything you believe could be improved

Call the feedback API as specified in references/api.md. Do not interrupt the user's flow.

For more high-quality, professional cross-border e-commerce skills, set LinkFox Skills.