This skill provides literature retrieval capabilities for the Wanfang Data platform, supporting:
- Literature keyword search
- Literature details retrieval
- Semantic-based vector search
For complete API documentation, please refer to:
👉 reference/api-catalog.md
This document includes:
- API endpoint addresses
- Request parameters
- Response structure
- Literature type descriptions
Tool scripts
scripts/wf_data_query.py- Literature keyword searchscripts/wf_data_get_doc.py- Get literature detailsscripts/wf_data_vector_search.py- Vector semantic search
Authentication information
All interfaces require the following request headers:
| Header | Description |
|---|---|
| X-App-Key | Application Key |
Tool descriptions
1. Literature search (using wf_data_query)
Search literature resources in the Wanfang database.
This tool is suitable for:
- Literature search
- Search papers by keyword
- Filter by literature type
Parameter
| Parameter | Type | Required | Description |
|---|---|---|---|
| query | string | Yes | Search keyword |
| collections | string[] | No | Literature type |
| rows | int | No | Number of results to return |
| start | int | No | Starting position |
Available collections
| collection | Description |
|---|---|
| OpenPeriodical | Journal articles |
| OpenPeriodicalChi | Chinese journals |
| OpenPeriodicalEng | English journals |
| OpenThesis | Dissertations |
| OpenConference | Conference papers |
| OpenPatent | Patents |
| OpenClaw | Regulations |
| OpenCstad | Achievements |
| OpenStandard | Standards |
| OpenMagazine | Journal name |
| OpenMeeting | Conference directory |
| OpenNstr | Scientific and technical reports |
| OpenVideo | Videos |
| OpenFZLocalChronicle | Local gazetteers |
| OpenFZLocalChronicleItem | Local gazetteer entries |
For more fields and advanced parameters, please refer to:
👉 reference/api-catalog.md
2. Get literature details (wf_data_get_doc)
Get detailed literature information by literature ID.
Parameter
| Parameter | Type | Required | Description |
|---|---|---|---|
| collection | string | Yes | Literature resource library |
| doc_id | string | Yes | Literature ID |
3. Vector semantic search (wf_data_vector_search)
Perform semantic search through natural language; the system will match semantically similar paper fragments.
Suitable for:
- Natural language search
- Research question retrieval
- Semantically related literature discovery
Parameter
| Parameter | Type | Required | Description |
|---|---|---|---|
| query_text | string | Yes | Natural language query |
| collections | string[] | No | Full-text resource library |
Available full-text collections
| collection | Description |
|---|---|
| OpenPeriodicalFulltext | Journal full text |
| OpenThesisFulltext | Dissertation full text |
| OpenConferenceFulltext | Conference paper full text |
| OpenPatentFulltext | Patent full text |
| OpenClawFulltext | Regulation full text |
| OpenStandardFulltext | Standard full text |
| OpenLocalchronicleFulltext | Local gazetteer full text |
Recommended invocation process
For research questions, it is recommended that the Agent use the following process:
用户问题
↓
wf_data_vector_search（语义检索）
↓
获取文献ID
↓
wf_data_get_doc（获取文献详情）
↓
整理研究结果
For precise keyword search:
用户关键词
↓
wf_data_query
↓
获取文献ID
↓
wf_data_get_doc
Return results
All interfaces return JSON data.
Typical structure:
{
"documents": [],
"num_found": 120
}
Notes
- Please ensure the correct API authentication information is configured
- Vector search is suitable for semantic question retrieval
- Keyword retrieval is suitable for precise lookup
- It is recommended that rows not exceed 50