Introducing Agents for Amazon Bedrock

Primer on Advanced Reasoning and ReAct

Create an Agent for Amazon Bedrock

1. Provide agent details including agent name, description (optional), whether the agent is allowed to request additional user inputs, and the AWS Identity and Access Management (IAM) service role that gives your agent access to other required services, such as Amazon Simple Storage Service (Amazon S3) and AWS Lambda.
提供代理详细信息，包括代理名称、描述（可选）、是否允许代理请求其他用户输入，以及使您的代理能够访问其他所需服务（例如 Amazon Simple Storage）的 AWS Identity and Access Management (IAM) 服务角色服务 (Amazon S3) 和 AWS Lambda。

2. Select a foundation model from Bedrock that fits your use case. Here, you provide an instruction to your agent in natural language. The instruction tells the agent what task it’s supposed to perform and the persona it’s supposed to assume. For example, “You are an agent designed to help with processing insurance claims and managing pending paperwork.”
从 Bedrock 中选择适合您的使用案例的基础模型。在这里，您可以用自然语言向您的代理提供指示。该指令告诉代理应该执行什么任务以及应该承担的角色。例如，“您是一名代理人，旨在帮助处理保险索赔和管理待处理的文书工作。”

3. Add action groups. An action is a task that the agent can perform automatically by making API calls to your company systems. A set of actions is defined in an action group. Here, you provide an API schema that defines the APIs for all the actions in the group. You also must provide a Lambda function that represents the business logic for each API. For example, let’s define an action group called ClaimManagementActionGroup that manages insurance claims by pulling a list of open claims, identifying outstanding paperwork for each claim, and sending reminders to policy holders. Make sure to capture this information in the action group description.The business logic for my action group is captured in the Lambda function InsuranceClaimsLambda. This AWS Lambda function implements methods for the following API calls: open-claims, identify-missing-documents, and send-reminders.Here’s a short extract from my OrderManagementLambda:
添加操作组。操作是代理可以通过对公司系统进行 API 调用来自动执行的任务。操作组中定义了一组操作。在这里，您提供一个 API 架构，该架构定义组中所有操作的 API。您还必须提供一个 Lambda 函数来表示每个 API 的业务逻辑。例如，让我们定义一个名为 ClaimManagementActionGroup 的操作组，该组通过提取未结索赔列表、识别每项索赔的未完成文书工作并向保单持有人发送提醒来管理保险索赔。确保在操作组描述中捕获此信息。



我的操作组的业务逻辑在 Lambda 函数 InsuranceClaimsLambda 中捕获。此 AWS Lambda 函数实现了以下 API 调用的方法： open-claims 、 identify-missing-documents 和 send-reminders. 下面是我的 OrderManagementLambda 的简短摘录：
Python Python
Note that you also must provide an API schema in the OpenAPI schema JSON format. Here’s what my API schema file insurance_claim_schema.json looks like:
请注意，您还必须提供 OpenAPI 架构 JSON 格式的 API 架构。我的 API 架构文件 insurance_claim_schema.json 如下所示：
JSON JSON
When a user asks your agent to complete a task, Bedrock will use the FM you configured for the agent to identify the sequence of actions and invoke the corresponding Lambda functions in the right order to solve the user-requested task.
当用户要求您的代理完成任务时，Bedrock 将使用您为代理配置的 FM 来识别操作顺序，并按正确的顺序调用相应的 Lambda 函数来解决用户请求的任务。

4. In the final step, review your agent configuration and choose Create Agent.
在最后一步中，检查您的代理配置并选择创建代理。

5. Congratulations, you’ve just created your first agent in Amazon Bedrock!
恭喜，您刚刚在 Amazon Bedrock 中创建了第一个代理！

Deploy an Agent for Amazon Bedrock

1. In the Bedrock console, select your agent, then select Deploy, and choose Create to create an alias.
在 Bedrock 控制台中，选择您的代理，然后选择部署，然后选择创建以创建别名。

2. Provide an alias name and description and choose whether to create a new version or use an existing version of your agent to associate with this alias.
提供别名和描述，并选择是创建新版本还是使用代理的现有版本与此别名关联。

3. This saves a snapshot of the agent code and configuration and associates an alias with this snapshot or version. You can use the alias to integrate the agent into your applications.
这将保存代理代码和配置的快照，并将别名与该快照或版本相关联。您可以使用别名将代理集成到您的应用程序中。