WorkingMemory

WorkingMemory 是一个输入处理器，用于将工作内存数据注入为系统消息。它从存储中检索持久信息，并将其格式化为对大语言模型的指令，使代理能够在对话中保持对用户的上下文理解。

🌐 The WorkingMemory is an input processor that injects working memory data as a system message. It retrieves persistent information from storage and formats it as instructions for the LLM, enabling the agent to maintain context about users across conversations.

使用示例

🌐 Usage example

import { WorkingMemory } from "@mastra/core/processors";

const processor = new WorkingMemory({
  storage: memoryStorage,
  scope: "resource",
  template: {
    format: "markdown",
    content: `# User Profile
- **Name**:
- **Preferences**:
- **Goals**:
`,
  },
});

构造函数参数

🌐 Constructor parameters

options:

Options

Configuration options for the working memory processor

选项

🌐 Options

storage:

MemoryStorage

Storage instance for retrieving working memory data

template?:

WorkingMemoryTemplate

Template defining the format and structure of working memory

scope?:

'thread' | 'resource'

Scope of working memory. 'thread' scopes to current thread, 'resource' shares across all threads for the resource

useVNext?:

boolean

Use the next-generation instruction format with improved guidelines

readOnly?:

boolean

When true, working memory is provided as read-only context. The data is injected into the conversation but without the updateWorkingMemory tool or update instructions. Useful for agents that should reference working memory without modifying it.

templateProvider?:

{ getWorkingMemoryTemplate(args: { memoryConfig?: MemoryConfig }): Promise<WorkingMemoryTemplate | null> }

Dynamic template provider for runtime template resolution

logger?:

IMastraLogger

Optional logger instance for structured logging

WorkingMemoryTemplate

format:

'markdown' | 'json'

Format of the working memory content

content:

string

Template content defining the structure of working memory data

返回

🌐 Returns

id:

string

Processor identifier set to 'working-memory'

name:

string

Processor display name set to 'WorkingMemory'

defaultWorkingMemoryTemplate:

string

The default markdown template used when no custom template is provided

processInput:

(args: { messages: MastraDBMessage[]; messageList: MessageList; abort: (reason?: string) => never; requestContext?: RequestContext }) => Promise<MessageList | MastraDBMessage[]>

Retrieves working memory and adds it as a system message to the message list

扩展使用示例

🌐 Extended usage example

src/mastra/agents/personalized-agent.ts

import { Agent } from "@mastra/core/agent";
import { WorkingMemory, MessageHistory } from "@mastra/core/processors";
import { PostgresStorage } from "@mastra/pg";

const storage = new PostgresStorage({
  connectionString: process.env.DATABASE_URL,
});

export const agent = new Agent({
  name: "personalized-agent",
  instructions: "You are a helpful assistant that remembers user preferences",
  model: "openai:gpt-4o",
  inputProcessors: [
    new WorkingMemory({
      storage,
      scope: "resource",
      template: {
        format: "markdown",
        content: `# User Information
- **Name**:
- **Location**:
- **Preferences**:
- **Communication Style**:
- **Current Projects**:
`,
      },
    }),
    new MessageHistory({ storage, lastMessages: 50 }),
  ],
  outputProcessors: [
    new MessageHistory({ storage }),
  ],
});

JSON 格式示例

🌐 JSON format example

import { WorkingMemory } from "@mastra/core/processors";

const processor = new WorkingMemory({
  storage: memoryStorage,
  scope: "resource",
  template: {
    format: "json",
    content: JSON.stringify({
      user: {
        name: { type: "string" },
        preferences: { type: "object" },
        goals: { type: "array" },
      },
    }),
  },
});

行为

🌐 Behavior

输入处理

🌐 Input processing

1. 从请求上下文中获取 threadId 和 resourceId
2. 根据作用域，从以下位置获取工作内存：
   • 线程元数据 (scope: 'thread')
   • 资源记录 (scope: 'resource')
3. 解析模板（来自提供商、选项或默认）
4. 根据模式生成系统指令：
   • 普通模式：包括存储/更新信息的指南、模板结构以及当前数据
   • 只读模式（readOnly: true）：仅包含当前数据作为上下文，不包含更新指令
5. 将该指令作为带有 source: 'memory' 标签的系统消息添加

工作内存更新

🌐 Working memory updates

工作内存的更新是通过 Memory 类提供的 updateWorkingMemory 工具完成的，而不是通过这个处理器。该处理器仅负责将当前的工作内存状态注入到对话中。

🌐 Working memory updates happen through the updateWorkingMemory tool provided by the Memory class, not through this processor. The processor only handles injecting the current working memory state into conversations.

默认模板

🌐 Default template

如果未提供模板，处理器将使用带有以下字段的默认 Markdown 模板：

🌐 If no template is provided, the processor uses a default markdown template with fields for:

• 名, 姓
• 地点，职业
• 兴趣，目标
• 事件、事实、项目

相关

🌐 Related

• 护栏