# Managing Threads and Messages

线索（Threads）和消息（Messages）构成了助手与用户之间的对话会话。Afarensis 助手API通过这两种对象，提供了一种灵活的方式来管理对话的内容和上下文。

创建线索和消息 创建线索：您可以在创建线索时包含一个初始消息列表。这允许您从特定的对话内容开始对话会话。

thread = client.beta.threads.create( messages=\[ { "role": "user", "content": "Create 3 data visualizations based on the trends in this file.", "file\_ids": \[file.id] } ] ) 消息内容：消息可以包含文本、图像或文件。目前，用户创建的消息不支持包含图像文件，但我们计划未来添加此功能。消息的文件大小和令牌限制与助手相同（512 MB 文件大小限制和 2,000,000 令牌限制）。

上下文窗口管理 Afarensis 助手API自动管理上下文窗口，以确保对话的内容不会超出模型的上下文长度限制。当线索中的消息内容超过模型的上下文窗口时，线索将尽可能包含适合上下文窗口的消息，并丢弃最旧的消息。

消息注释 注释类型：消息创建的注释有两种类型：file\_citation 和 file\_path。file\_citation 注释由检索工具创建，定义了对特定文件中特定引文的引用；file\_path 注释由代码解释器工具创建，包含对工具生成的文件的引用。

处理注释：当消息对象中存在注释时，您可能会看到文本中的不清楚的模型生成的子字符串。以下是一个示例代码片段，展示了如何用注释中的信息替换这些字符串：

## Retrieve the message object

message = client.beta.threads.messages.retrieve( thread\_id="...", message\_id="..." )

## Extract and process message content with annotations

message\_content = message.content\[0].text annotations = message\_content.annotations citations = \[]

for index, annotation in enumerate(annotations): message\_content.value = message\_content.value.replace(annotation.text, f' \[{index}]') if (file\_citation := getattr(annotation, 'file\_citation', None)): cited\_file = client.files.retrieve(file\_citation.file\_id) citations.append(f'\[{index}] {file\_citation.quote} from {cited\_file.filename}') elif (file\_path := getattr(annotation, 'file\_path', None)): cited\_file = client.files.retrieve(file\_path.file\_id) citations.append(f'\[{index}] Click to download {cited\_file.filename}')

message\_content.value += '\n' + '\n'.join(citations)


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://doc.afarensis.com/google-zhu-li-de-gong-zuo-yuan-li/managing-threads-and-messages.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.