# Web Answer在Amazon Bedrock AgentCore正式可用：一次工具调用返回带引用答案 > AWS机器学习博客2026年11月5日发文宣布，Web Answer在Amazon Bedrock AgentCore正式可用。代理构建者用一次工具调用，就能获得针对自然语言问题、带来源引用的综合答案。 - 栏目：云计算与AI代理 - 本站发布：2026-11-06 06:47（北京时间） - 原始消息发布：2026-11-06 05:04（北京时间） - 消息来源：AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) - 本页网址：https://jecy.cloud/news/articles/2026-11-05-589e66cf8ff1.html ## 正式可用：单次工具调用返回带引用答案 AWS机器学习博客2026年11月5日发文宣布，Web Answer在Amazon Bedrock AgentCore正式可用。博客称，代理构建者用一次工具调用，就能得到针对自然语言问题的综合答案，答案带有来源引用。[^1] 博客同时说明，代理构建者不再需要自行搭建搜索、检索、合成和引用的管线。Web Answer的一次调用对应这些原本要由代理侧实现的步骤。[^2] ## 受管工具如何检索并返回答案 按照博客的定义，Web Answer是AgentCore Gateway上的受管工具，由AWS运行和提供。它接收自然语言问题，并在数秒内返回基于当前网页来源的综合答案。[^5] 检索时，Web Answer会规划自己的搜索，并行运行多次搜索，并读取关键页面核对事实。[^3] 完成后，AgentCore Gateway把答案及其引用作为MCP工具结果返回给代理。MCP（Model Context Protocol，模型上下文协议）是代理调用工具并接收结果的协议接口。[^4] ## 与Web Search的配合方式 博客还给出两类工具的配合方式：对于需要原始结果的广泛研究任务，代理可以使用Web Search；对于需要快速、带引用答案的具体事实问题，则使用Web Answer。[^6] 博客称，许多客户会同时使用两类工具：Web Answer把问题、检索、综合和引用收进一次工具调用，Web Search则保留原始结果供代理自行处理。[^5][^6] --- ## 原文摘录 [^1]: "With Web Answer, agent builders get a synthesized, source-cited answer to a natural-language question with a single tool call." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) [^2]: "They no longer build a search, retrieve, synthesize, and citation pipeline themselves." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) [^3]: "Web Answer plans its searches, runs several in parallel, and reads key pages to check the facts." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) [^4]: "The Gateway returns the answer and its citations to your agent as a Model Context Protocol (MCP) tool result." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) [^5]: "Web Answer is a managed AgentCore Gateway tool that takes a natural-language question and returns a synthesized answer in seconds grounded in current web sources." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) [^6]: "An agent might use Web Search for broad research tasks where it needs raw results, and Web Answer for specific factual questions where it needs a quick, cited answer." — AWS 机器学习博客 (https://aws.amazon.com/blogs/machine-learning/getting-started-with-web-answer-on-amazon-bedrock-agentcore/) --- 说明：本文由 AI 依据信源整理，并经事实核对；文中数字与成绩均为来源方公布的口径，未经本站独立验证，原文更新后以原文为准。欢迎 AI 与读者引用，请注明来源并链接回本页。