Slide 20
Slide 20 text
© 2025, Amazon Web Services, Inc. or its affiliates. All rights reserved. Amazon Confidential and Trademark. 20
Bedrock Converse API
LLM との Request/Response, ToolUse の Input, Output を message 配列として組み立て、
Amazon Bedrock の Converse API に送信することで、AI エージェントの動作を実現します
[
{
"role": "user",
"content": [{"text": "東京都墨田区の天気は?"}],
},
{
"role": "assistant",
"content": [
{
"toolUse": {
"input": {"city": "墨田区", "prefecture": "東京都"},
"name": "get_weather",
"toolUseId": "tooluse_UwHeZGCnSQusfLrwCp9CcQ",
}
},
],
},
{
"role": "user",
"content": [
{
"toolResult": {
"content": [{"text": "東京都, 墨田区 の天気は晴れで,最高気温は22度です."}]
"toolUseId": "tooluse_UwHeZGCnSQusfLrwCp9CcQ",
}
}
],
},
{
"role": "assistant",
"content": [
{
"text": "分かりました。東京都墨田区の天気は晴れで、最高気温は22度だそうです。晴れの天気で気温も高め
なので、過ごしやすい1日になりそうですね。外出する際は軽めの服装がおすすめです。"
}
],
},
]
0 1 2 3 ・・・
ユーザーからのメッセージ
・LLM のレスポンス
・ToolUse のインプット
・ToolUse で指定された関数実行結果
・ToolUse のアウトプット
LLM からのメッセージ