Procurize.AI REST API 参考#

使用您的 API 令牌以编程方式访问您的知识库——文档、问卷、RAG 和报告。

适用范围： Procurize.AI API v1
身份验证： 通过 X-API-Key 请求头传递 API 密钥（参见 API 令牌 进行令牌管理）
基础 URL： https://api.procurize.com

概览#

Procurize.AI REST API 提供对平台关键功能的编程访问。它使您能够将 Procurize.AI 与内部系统集成、自动化知识库管理并构建自定义工作流。

受支持的模块（需令牌认证）：

身份验证#

所有需要身份验证的端点都必须在请求头中传递 API 令牌。

HeaderValue
X-API-Key<YOUR_API_TOKEN>

示例：

curl -X GET "https://api.procurize.com/v1/..." \
  -H "X-API-Key: your-api-token-here"

💡 令牌绑定到您的组织——无需在请求中指定组织 ID。

基础 URL#

所有 REST API 端点均位于以下地址下：

https://api.procurize.com

错误处理#

API 使用标准的 HTTP 状态码。

CodeMeaning
200 OK请求成功
201 Created资源创建成功
204 No Content请求成功，无响应体
400 Bad Request参数无效或请求格式错误
401 Unauthorized缺少或无效的 API 令牌
403 Forbidden令牌没有执行此操作的权限
404 Not Found资源不存在
500 Internal Server Error服务器内部错误

错误响应包含机器可读的错误代码和人类可读的错误信息。

模块参考#

RagArticle#

GET /questionnaire/rag/articles/summary#

摘要： 首页摘要

参数:

NameInTypeRequiredDescription
projquerystring (uuid)No项目 ID。

响应 (200):

{
  "active": 5
}

GET /questionnaire/rag/articles/total#

摘要： 文章数量

参数:

NameInTypeRequiredDescription
qquerystringNo按文章名称或代码过滤。
statequeryarrayNo按状态过滤。
projectqueryarrayNo按项目过滤。
typequeryarrayNo按文档类型过滤。

响应 (200):

42

GET /questionnaire/rag/articles/list#

摘要： 获取文章列表

参数:

NameInTypeRequiredDescription
qquerystringNo按文章名称或代码过滤。
statequeryarrayNo按状态过滤。
projectqueryarrayNo按项目过滤。
typequeryarrayNo按文档类型过滤。
squerystringNo排序字段。
ascquerybooleanNo升序排序。
psqueryinteger (int32)No每页大小。
pqueryinteger (int32)No基于 1 的页码。

响应 (200):

{
  "paging": {
    "pageNumber": 1,
    "pageSize": 10,
    "total": 42,
    "totalPages": 5
  },
  "items": [
    {
      "code": "ART-001",
      "org": "00000000-0000-0000-0000-000000000000",
      "id": "00000000-0000-0000-0000-000000000000",
      "name": "Article Title",
      "description": "Article description",
      "documentTypes": [
        {
          "id": "00000000-0000-0000-0000-000000000000",
          "name": "Type Name",
          "plural": "Type Names",
          "count": 5
        }
      ],
      "size": 150,
      "revision": {
        "id": "00000000-0000-0000-0000-000000000000",
        "collectionId": "00000000-0000-0000-0000-000000000000",
        "rev": 3,
        "date": "2025-01-15T10:30:00Z",
        "author": {
          "id": "user123",
          "name": "John Doe",
          "email": "john.doe@example.com",
          "login": "johndoe",
          "ra": true
        },
        "authorName": "John Doe",
        "analysisSummary": {
          "id": "00000000-0000-0000-0000-000000000000",
          "initiator": {
            "id": "user123",
            "name": "John Doe",
            "email": "john.doe@example.com",
            "login": "johndoe",
            "ra": true
          },
          "created": "2025-01-15T10:30:00Z",
          "state": "Completed",
          "stateDate": "2025-01-15T10:35:00Z",
          "stateDetails": "Analysis completed successfully",
          "operationId": "op-12345"
        }
      },
      "type": "Article",
      "state": "Active",
      "projects": [
        "00000000-0000-0000-0000-000000000000"
      ],
      "items": [
        "00000000-0000-0000-0000-000000000000"
      ],
      "isShared": false
    }
  ]
}

POST /questionnaire/rag/file-to-md#

摘要： 从文件返回 Markdown 内容

请求体 (multipart/form-data):

{
  "url": "https://example.com/file.pdf",
  "file": "binary"
}

响应 (200):

"Markdown content from file"

POST /questionnaire/rag/article/create#

摘要： 创建文章

请求体 (multipart/form-data):

{
  "code": "ART-001",
  "name": "Article Title",
  "description": "Article description",
  "content": "# Article Content\n\nThis is the article content.",
  "url": "https://example.com/file.pdf",
  "file": "binary"
}

响应 (200):

{
  "code": "ART-001",
  "org": "00000000-0000-0000-0000-000000000000",
  "id": "00000000-0000-0000-0000-000000000000",
  "name": "Article Title",
  "description": "Article description",
  "documentTypes": null,
  "size": 150,
  "revision": null,
  "type": "Article",
  "state": "Active",
  "projects": null,
  "items": null,
  "isShared": false
}

GET /questionnaire/rag/article/check#

摘要： 检查文章是否存在

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

响应 (200):

true

GET /questionnaire/rag/article-info#

摘要： 按代码获取文章信息

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

响应 (200):

{
  "code": "ART-001",
  "org": "00000000-0000-0000-0000-000000000000",
  "id": "00000000-0000-0000-0000-000000000000",
  "name": "Article Title",
  "description": "Article description",
  "documentTypes": null,
  "size": 150,
  "revision": null,
  "type": "Article",
  "state": "Active",
  "projects": null,
  "items": null,
  "isShared": false
}

GET /questionnaire/rag/article#

摘要： 按代码获取文章内容

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。
formatquerystringNo为 true 时返回 HTML 内容；否则返回 Markdown 格式的文章内容。
splitquerybooleanNo为 true 时返回按 RAG 元素分割的内容。
cssquerybooleanNo为 true 时返回带 CSS 的 HTML/PDF 内容。

响应 (200):

"# Article Title\n\nThis is the article content in Markdown format."

POST /questionnaire/rag/article#

摘要： 更新文章元数据

请求体:

{
  "code": "ART-001",
  "name": "Updated Article Title",
  "description": "Updated description",
  "projects": [
    "00000000-0000-0000-0000-000000000000"
  ],
  "documentTypes": [
    {
      "id": "00000000-0000-0000-0000-000000000000",
      "name": "Type Name",
      "plural": "Type Names"
    }
  ],
  "isShared": true,
  "state": "Active"
}

响应 (200):

{}

DELETE /questionnaire/rag/article#

摘要： 删除文章

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

GET /questionnaire/rag/article/pdf#

摘要： 按代码获取 PDF 形式的文章

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

响应 (200):

(pdf binary)

GET /questionnaire/rag/article/content#

摘要： 获取文章内容

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

响应 (200):

{
  "content": "# Article Title\n\nContent here."
}

POST /questionnaire/rag/article/content#

摘要： 更新文章内容

请求体 (multipart/form-data):

{
  "code": "ART-001",
  "content": "# Updated Content",
  "file": "binary"
}

POST /questionnaire/rag/plural#

摘要： 获取名词的复数形式

参数:

NameInTypeRequiredDescription
nounquerystringNo需要复数化的名词。

响应 (200):

{
  "plural": "Articles"
}

POST /questionnaire/rag/article/find-types#

摘要： 使用 AI 分析文章并返回其类型

请求体 (multipart/form-data):

{
  "code": "ART-001",
  "content": "# Article Content"
}

响应 (200):

[
  {
    "plural": "Types"
  }
]

POST /questionnaire/rag/article/format-with-ai#

摘要： 使用 AI 格式化文章内容

请求体 (multipart/form-data):

{
  "code": "ART-001",
  "content": "# Article Content"
}

响应 (200):

{
  "formattedContent": "# Formatted Article Content"
}

POST /questionnaire/rag/article/improve-on-suggestions#

摘要： 根据建议改进文档

请求体 (multipart/form-data):

{
  "suggestions": [
    {
      "sectionTitle": "Introduction",
      "suggestion": "Add more details here."
    }
  ],
  "code": "ART-001",
  "content": "# Article Content"
}

响应 (200):

{
  "improvedContent": "# Improved Content"
}

GET /questionnaire/rag/article/parts#

摘要： 返回文章部件列表

参数:

NameInTypeRequiredDescription
idqueryarrayYes文章部件列表。

响应 (200):

[
  {
    "id": "00000000-0000-0000-0000-000000000000",
    "code": "part-1",
    "title": "Introduction",
    "orderNumber": 1,
    "content": "Introduction content."
  }
]

GET /questionnaire/rag/article/revisions#

摘要： 返回文章修订列表

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。

响应 (200):

[
  {
    "id": "00000000-0000-0000-0000-000000000000",
    "collectionId": "00000000-0000-0000-0000-000000000000",
    "rev": 1,
    "date": "2025-01-15T10:30:00Z",
    "author": {
      "id": "user123",
      "name": "John Doe",
      "email": "john.doe@example.com",
      "login": "johndoe",
      "ra": true
    },
    "authorName": "John Doe",
    "analysisSummary": {
      "id": "00000000-0000-0000-0000-000000000000",
      "initiator": {
        "id": "user123",
        "name": "John Doe",
        "email": "john.doe@example.com",
        "login": "johndoe",
        "ra": true
      },
      "created": "2025-01-15T10:30:00Z",
      "state": "Completed",
      "stateDate": "2025-01-15T10:35:00Z",
      "stateDetails": "Analysis completed",
      "operationId": "op-12345"
    }
  }
]

GET /questionnaire/rag/article/revisions/info/{id}#

摘要： 获取文章修订信息

参数:

NameInTypeRequiredDescription
idpathstring (uuid)Yes修订 ID。

响应 (200):

{
  "id": "00000000-0000-0000-0000-000000000000",
  "collectionId": "00000000-0000-0000-0000-000000000000",
  "rev": 1,
  "date": "2025-01-15T10:30:00Z",
  "author": {
    "id": "user123",
    "name": "John Doe",
    "email": "john.doe@example.com",
    "login": "johndoe",
    "ra": true
  },
  "authorName": "John Doe",
  "analysisSummary": {
    "id": "00000000-0000-0000-0000-000000000000",
    "initiator": {
      "id": "user123",
      "name": "John Doe",
      "email": "john.doe@example.com",
      "login": "johndoe",
      "ra": true
    },
    "created": "2025-01-15T10:30:00Z",
    "state": "Completed",
    "stateDate": "2025-01-15T10:35:00Z",
    "stateDetails": "Analysis completed",
    "operationId": "op-12345"
  }
}

GET /questionnaire/rag/article/revisions/{id}#

摘要： 获取文章修订内容

参数:

NameInTypeRequiredDescription
idpathstring (uuid)Yes修订 ID。

响应 (200):

"# Article Revision Content\n\nPrevious version."

GET /questionnaire/rag/article/analyze#

摘要： 使用 AI 分析文章内容

参数:

NameInTypeRequiredDescription
codequerystringYes文章代码。
revqueryinteger (int32)No修订号。
forcequerybooleanNo强制创建新分析并删除已有的未完成分析结果。

响应 (200):

{
  "id": "00000000-0000-0000-0000-000000000000",
  "initiator": {
    "id": "user123",
    "name": "John Doe",
    "email": "john.doe@example.com",
    "login": "johndoe",
    "ra": true
  },
  "created": "2025-01-15T10:30:00Z",
  "state": "Completed",
  "stateDate": "2025-01-15T10:35:00Z",
  "stateDetails": "Analysis completed",
  "operationId": "op-12345"
}

GET /questionnaire/rag/article/analyze/{id}#

摘要： 获取文章分析结果

参数:

NameInTypeRequiredDescription
idpathstring (uuid)Yes分析 ID。

响应 (200):

{
  "id": "00000000-0000-0000-0000-000000000000",
  "initiator": {
    "id": "user123",
    "name": "John Doe",
    "email": "john.doe@example.com",
    "login": "johndoe",
    "ra": true
  },
  "created": "2025-01-15T10:30:00Z",
  "state": "Completed",
  "stateDate": "2025-01-15T10:35:00Z",
  "stateDetails": "Analysis completed",
  "operationId": "op-12345",
  "fragments": [
    "fragment-1",
    "fragment-2"
  ],
  "ownInconcies": "Own inconsistencies found",
  "ragInconcies": "RAG inconsistencies found",
  "proposals": "Improvement proposals"
}

GET /questionnaire/rag/messages/list#

摘要： 列出所有消息

参数:

NameInTypeRequiredDescription
idquerystring (uuid)YesRAG 集合 ID。
itemquerystring (uuid)NoRAG 项目 ID。
messagequerystring (uuid)No消息 ID。
ascquerybooleanNo升序排序。
psqueryinteger (int32)No每页大小。
pqueryinteger (int32)No基于 1 的页码。

响应 (200):

{
  "paging": {
    "pageNumber": 1,
    "pageSize": 10,
    "total": 5,
    "totalPages": 1
  },
  "items": [
    {
      "id": "00000000-0000-0000-0000-000000000000",
      "replyTo": null,
      "questionnaire": null,
      "section": null,
      "question": null,
      "ragCollection": "00000000-0000-0000-0000-000000000000",
      "ragItem": "00000000-0000-0000-0000-000000000000",
      "targetId": "00000000-0000-0000-0000-000000000000",
      "targetType": "Question",
      "targetNum": 1,
      "author": {
        "id": "user123",
        "name": "John Doe",
        "email": "john.doe@example.com",
        "login": "johndoe",
        "ra": true
      },
      "text": "Message content",
      "created": "2025-01-15T10:30:00Z",
      "updated": "2025-01-15T10:30:00Z",
      "replies": 0
    }
  ]
}

POST /questionnaire/rag/messages/add#

摘要： 添加新消息

参数:

NameInTypeRequiredDescription
idquerystring (uuid)YesRAG 集合 ID。
itemquerystring (uuid)NoRAG 项目 ID。
messagequerystring (uuid)No要回复的消息 ID。

请求体:

{
  "text": "New message content"
}

响应 (200):

{
  "id": "00000000-0000-0000-0000-000000000000",
  "replyTo": null,
  "questionnaire": null,
  "section": null,
  "question": null,
  "ragCollection": "00000000-0000-0000-0000-000000000000",
  "ragItem": "00000000-0000-0000-0000-000000000000",
  "targetId": "00000000-0000-0000-0000-000000000000",
  "targetType": "Question",
  "targetNum": 1,
  "author": {
    "id": "user123",
    "name": "John Doe",
    "email": "john.doe@example.com",
    "login": "johndoe",
    "ra": true
  },
  "text": "New message content",
  "created": "2025-01-15T10:30:00Z",
  "updated": "2025-01-15T10:30:00Z",
  "replies": 0
}

POST /questionnaire/rag/message#

摘要： 更新已有消息

参数:

NameInTypeRequiredDescription
cquerystring (uuid)YesRAG 集合 ID。
idquerystring (uuid)Yes消息 ID。

请求体:

{
  "text": "Updated message content"
}

响应 (200):

{
  "id": "00000000-0000-0000-0000-000000000000",
  "replyTo": null,
  "questionnaire": null,
  "section": null,
  "question": null,
  "ragCollection": "00000000-0000-0000-0000-000000000000",
  "ragItem": "00000000-0000-0000-0000-000000000000",
  "targetId": "00000000-0000-0000-0000-000000000000",
  "targetType": "Question",
  "targetNum": 1,
  "author": {
    "id": "user123",
    "name": "John Doe",
    "email": "john.doe@example.com",
    "login": "johndoe",
    "ra": true
  },
  "text": "Updated message content",
  "created": "2025-01-15T10:30:00Z",
  "updated": "2025-01-15T10:45:00Z",
  "replies": 0
}

DELETE /questionnaire/rag/message#

摘要： 删除已有消息

参数:

NameInTypeRequiredDescription
idquerystring (uuid)Yes消息 ID。

响应 (200):

{}

RagFaq#

GET /questionnaire/rag/faq-collections/summary#

摘要： 首页摘要

参数:

NameInTypeRequiredDescription
projquerystring (uuid)No项目 ID。

响应 (200):

{
  "active": 3
}

GET /questionnaire/rag/faq-collections/total#

摘要： FAQ 集合数量

参数:

NameInTypeRequiredDescription
qquerystringNo按集合名称或代码过滤。
statequeryarrayNo按状态过滤。
projectqueryarrayNo按项目过滤。

响应 (200):

3

GET /questionnaire/rag/faq-collections/list#

摘要： 获取 FAQ 集合列表

参数:

NameInTypeRequiredDescription
qquerystringNo按集合名称或代码过滤。
statequeryarrayNo按状态过滤。
projectqueryarrayNo按项目过滤。
squerystringNo排序字段。
ascquerybooleanNo升序排序。
psqueryinteger (int32)No每页大小。
pqueryinteger (int32)No基于 1 的页码。

响应 (200):

{
  "paging": {
    "pageNumber": 1,
    "pageSize": 10,
    "total": 3,
    "totalPages": 1
  },
  "items": [
    {
      "code": "FAQ-001",
      "org": "00000000-0000-0000-0000-000000000000",
      "id": "00000000-0000-0000-0000-000000000000",
      "name": "FAQ Collection",
      "description": "FAQ collection description",
      "documentTypes": null,
      "size": 5,
      "revision": null,
      "type": "Faq",
      "state": "Active",
      "projects": null,
      "items": null,
      "isShared": false
    }
  ]
}

GET /questionnaire/rag/faq-collections/find-items#

摘要： 查找 FAQ 条目

参数:

NameInTypeRequiredDescription
qquerystringYes按问题过滤。
projectquerystring (uuid)No按项目过滤。
psqueryinteger (int32)No每页大小。
pqueryinteger (int32)No基于 1 的页码。

响应 (200):

{
  "paging": {
    "pageNumber": 1,
    "pageSize": 10,
    "total": 5,
    "totalPages": 1
  },
  "items": [
    {
      "index": 0,
      "question": "What is this product?",
      "answer": "This is a great product.",
      "comment": "Product FAQ",
      "state": "Active"
    }
  ]
}

POST /questionnaire/rag/faq-collection/create#

摘要： 创建新的 FAQ 集合

请求体 (multipart/form-data):

{
  "code": "FAQ-001",
  "name": "FAQ Collection",
  "description": "FAQ collection description",
  "content": "Content here",
  "file": "binary"
}

响应 (200):

{
  "code": "FAQ-001",
  "id": "00000000-0000-0000-0000-000000000000",
  "name": "FAQ Collection",
  "description": "FAQ collection description",
  "documentTypes": null,
  "size": 5,
  "revision": null,
  "type": "Faq",
  "state": "Active",
  "projects": null,
  "items": null,
  "isShared": false
}

POST /questionnaire/rag/faq-collection#

摘要： 更新 FAQ 集合元数据

请求体:

{
  "code": "FAQ-001",
  "name": "Updated FAQ Collection",
  "description": "Updated description",
  "projects": null,
  "documentTypes": null,
  "isShared": false,
  "state": "Active"
}

DELETE /questionnaire/rag/faq-collection#

摘要： 删除 FAQ 集合

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。

GET /questionnaire/rag/faq-collection#

摘要： 获取 FAQ 集合

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。

响应 (200):

{
  "code": "FAQ-001",
  "org": "00000000-0000-0000-0000-000000000000",
  "id": "00000000-0000-0000-0000-000000000000",
  "name": "FAQ Collection",
  "description": "FAQ collection description",
  "documentTypes": null,
  "size": 5,
  "revision": null,
  "type": "Faq",
  "state": "Active",
  "projects": null,
  "items": null,
  "isShared": false
}

GET /questionnaire/rag/faq-collection/check#

摘要： 检查 FAQ 集合是否存在

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。

响应 (200):

true

GET /questionnaire/rag/faq-collection/items-total#

摘要： FAQ 条目数量

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。
qquerystringNo按 FAQ 条目内容过滤。

请求体:

["Active", "Default"]

响应 (200):

5

GET /questionnaire/rag/faq-collection/items#

摘要： 获取 FAQ 条目列表

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。
qquerystringNo按 FAQ 条目内容过滤。
statequeryarrayNo按状态过滤。
squerystringNo排序字段。
ascquerybooleanNo升序排序。
psqueryinteger (int32)No每页大小。
pqueryinteger (int32)No基于 1 的页码。

响应 (200):

{
  "paging": {
    "pageNumber": 1,
    "pageSize": 10,
    "total": 5,
    "totalPages": 1
  },
  "items": [
    {
      "index": 0,
      "question": "What is this product?",
      "answer": "This is a great product.",
      "comment": "Product FAQ",
      "state": "Active"
    }
  ]
}

POST /questionnaire/rag/faq-collection/item/create#

摘要： 创建新的 FAQ 条目

请求体:

{
  "code": "FAQ-001",
  "index": 0,
  "question": "What is this product?",
  "answer": "This is a great product.",
  "comment": "Product FAQ",
  "state": "Active"
}

响应 (200):

{
  "index": 0,
  "question": "What is this product?",
  "answer": "This is a great product.",
  "comment": "Product FAQ",
  "state": "Active"
}

GET /questionnaire/rag/faq-collection/item#

摘要： 获取 FAQ 条目

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。
indexqueryinteger (int32)Yes集合中 FAQ 条目的零基索引。

响应 (200):

{
  "index": 0,
  "question": "What is this product?",
  "answer": "This is a great product.",
  "comment": "Product FAQ",
  "state": "Active"
}

POST /questionnaire/rag/faq-collection/item#

摘要： 更新 FAQ 条目

请求体:

{
  "code": "FAQ-001",
  "index": 0,
  "newIndex": 1,
  "question": "Updated question?",
  "answer": "Updated answer.",
  "comment": "Updated comment",
  "state": "Active"
}

DELETE /questionnaire/rag/faq-collection/item#

摘要： 删除 FAQ 条目

参数:

NameInTypeRequiredDescription
codequerystringYes集合代码。
indexqueryinteger (int32)Yes集合中 FAQ 条目的零基索引。

RagQuery#

POST /questionnaire/rag/query#

摘要： 在 RAG 知识库中执行查询

请求体:

{
  "project": "00000000-0000-0000-0000-000000000000",
  "excludeCollections": [
    "00000000-0000-0000-0000-000000000000"
  ],
  "question": "What is this product?",
  "yourCompanyName": "Acme Corp",
  "yourCompanyAlias": "Acme",
  "customerName": "John Doe",
  "options": [
    "option1",
    "option2"
  ],
  "multiselect": false,
  "nArticles": 5,
  "nFaqs": 3,
  "threshold": 0.5,
  "debug": false,
  "withJustification": true,
  "withConfidenceLevel": true,
  "withFragments": true,
  "nAttempts": 3
}

响应 (200):

{
  "attempt": 1,
  "preparedQuery": {
    "original": "What is this product?",
    "enhanced": "What is this product and its features?"
  },
  "ragItems": [
    {
      "id": "00000000-0000-0000-0000-000000000000",
      "distance": 0.25
    }
  ],
  "question": "What is this product?",
  "generalInstructions": "Based on the knowledge base, here is the answer:",
  "text": "This product is a great solution for your needs.",
  "explanation": "The answer was found in multiple sources with high confidence."
}

POST /questionnaire/rag/query-stream#

摘要： 在 RAG 知识库中执行流式查询

请求体:

{
  "project": "00000000-0000-0000-0000-000000000000",
  "excludeCollections": [
    "00000000-0000-0000-0000-000000000000"
  ],
  "question": "What is this product?",
  "yourCompanyName": "Acme Corp",
  "yourCompanyAlias": "Acme",
  "customerName": "John Doe",
  "options": [
    "option1",
    "option2"
  ],
  "multiselect": false,
  "nArticles": 5,
  "nFaqs": 3,
  "threshold": 0.5,
  "debug": false,
  "withJustification": true,
  "withConfidenceLevel": true,
  "withFragments": true,
  "nAttempts": 3
}

响应 (200):

[
  {
    "text": "This product is a great solution",
    "fragments": [
      {
        "id": "00000000-0000-0000-0000-000000000000",
        "part": 1,
        "distance": 0.25,
        "type": "Article",
        "code": "ART-001",
        "title": "Product Title",
        "fragment": "Product content fragment"
      }
    ]
  },
  {
    "text": " for your needs.",
    "fragments": []
  }
]

POST /questionnaire/rag/query-fragments#

摘要： 在 RAG 知识库中查询片段

请求体:

{
  "project": "00000000-0000-0000-0000-000000000000",
  "exclude_documents": [
    "00000000-0000-0000-0000-000000000000"
  ],
  "query": "What is this product?",
  "top_k": 5,
  "similarity_threshold": 0.5,
  "prepare_query": true
}

响应 (200):

{
  "results": [
    {
      "document_id": "00000000-0000-0000-0000-000000000000",
      "document_code": "ART-001",
      "document_title": "Product Title",
      "revision": 3,
      "chunk_index": 0,
      "content": "Product content fragment",
      "metadata": {
        "author": "John Doe",
        "created_at": "2025-01-15T10:30:00Z",
        "last_updated_at": "2025-01-15T10:35:00Z",
        "doc_categories": [
          "Product",
          "Information"
        ]
      },
      "similarity_score": 0.85
    }
  ]
}

相关文档#

到顶部
选择语言
English
Български
Čeština
Dansk
Deutsch
Ελληνική
Eestlane
Español
Suomalainen
Français
Hrvatski
Magyar
Հայերեն
Indonesia
Italiano
Lietuvių
Melayu
Nederlands
Polski
Português
Română
Русский
Slovenský
Svenska
ไทย
Türk
Українська
Tiếng Việt
한국어
日本語
中文
العربية
ქართული
עִברִית
हिंदी
فارسی