Merge pull request #2864 from kakakakakku/apigw-rest-stepfunctions-express-sync-bedrock-sam

julianwood · web-flow · commit a4314e6b17b6 · 2026-04-24T17:27:07.000+01:00
apigw-rest-stepfunctions-express-sync-bedrock-sam: Migrate from Cohere Command R to Amazon Nova 2 Lite
diff --git a/apigw-rest-stepfunctions-express-sync-bedrock-sam/Readme.md b/apigw-rest-stepfunctions-express-sync-bedrock-sam/Readme.md
@@ -1,6 +1,6 @@
 # Prompt Chaining with Amazon API Gateway, AWS Step Functions and Amazon Bedrock.
 
-The AWS Serverless Application Model (SAM) template deploys an Amazon API Gateway HTTP API endpoint connected to an AWS Step Functions state machine. This example demonstrates how to invoke an Express state machine synchronously and utilize AWS Step Functions intrinsic functions to chain two prompts, which are then used to invoke the Amazon Bedrock language model. The output from the state machine execution is returned to the client within 29 seconds, using the HTTP API. This no-code example showcases how the results from the first prompt can be used to provide context for the second prompt, allowing the language model to deliver a highly-curated response. By chaining these prompts, the system can leverage the capabilities of the language model to generate more meaningful and contextual outputs.
+The AWS Serverless Application Model (SAM) template deploys an Amazon API Gateway REST API endpoint connected to an AWS Step Functions state machine. This example demonstrates how to invoke an Express state machine synchronously and utilize AWS Step Functions intrinsic functions to chain two prompts, which are then used to invoke the Amazon Bedrock language model. The output from the state machine execution is returned to the client within 29 seconds, using the REST API. This no-code example showcases how the results from the first prompt can be used to provide context for the second prompt, allowing the language model to deliver a highly-curated response. By chaining these prompts, the system can leverage the capabilities of the language model to generate more meaningful and contextual outputs.
 
 Learn more about this pattern at [Serverless Land Patterns](https://serverlessland.com/patterns/apigw-rest-stepfunctions-express-sync-bedrock-sam)
 
@@ -12,7 +12,7 @@ Important: this application uses various AWS services and there are costs associ
 * [AWS CLI](https://docs.aws.amazon.com/cli/latest/userguide/install-cliv2.html) installed and configured
 * [Git Installed](https://git-scm.com/book/en/v2/Getting-Started-Installing-Git)
 * [AWS Serverless Application Model](https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/serverless-sam-cli-install.html) (AWS SAM) installed
-* [NOTE! Manage Access to Amazon Bedrock Foundation Models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-access.html) at the time of writing, this example uses Amazon Bedrock foundation model cohere.command-text-v14
+* This example uses Amazon Bedrock foundation model amazon.nova-2-lite-v1:0
 
 
 ## Deployment Instructions
@@ -45,40 +45,33 @@ Important: this application uses various AWS services and there are costs associ
 
 ## How it Works
 In this example, the state machine is invoked with a JSON payload
+
 ```asl
 {
   "prompt_one": "Write a 500 word blog post on The Beatles"
 }
 ```
+
 During execution, the Task state calls the Bedrock API and the response is passed to the task 'result_one'.
+
 ```asl
 {
-  "result_one.$": "$.Body.generations[0].text"
-}
-```
-A Pass state is then used to format the data using an Intrinsic Function (States.) which is passed to the next state.
-```asl
-{
-  "convo_one.$": "States.Format('{}',$.result_one.result_one)"
-}
-```
-The second prompt is then executed with new instructions and the results from the first execution. This provides the prompt with more context
-```asl
-{
-  "prompt.$": "States.Format('{}\n{}','Human: Now write a short story based on the following. Assistant:', $.convo_one.convo_one)",
-  "max_tokens": 1000
+  "result_one.$": "$.Body.output.message.content[0].text"
 }
 ```
-By default, the state then sends the task result as output.
 
+A Pass state is then used to build the conversation history with the first prompt, the model's response, and the second prompt, which is passed to the next state.
+
+The second prompt is then executed with the conversation history that includes the results from the first execution. This provides the model with more context.
+
+By default, the state then sends the task result as output.
 
 ## Testing
 
 The stack will output the **api endpoint**. You can use *Postman* or *curl* to send a POST request to the API Gateway endpoint.
    
 ```
 curl -H "Content-type: application/json" -X POST -d '{"prompt_one": "Write a 500 word blog post on The Beatles"}' <Your API endpoint>
-
 ```
 After runnning the above command, API Gateway will invoke the State machine and return the results back to the client instead of just the State machine's execution Id. 
 
@@ -93,6 +86,6 @@ After runnning the above command, API Gateway will invoke the State machine and
     aws cloudformation list-stacks --query "StackSummaries[?contains(StackName,'STACK_NAME')].StackStatus"
     ```
 ----
-Copyright 2024 Amazon.com, Inc. or its affiliates. All Rights Reserved.
+Copyright 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
 
-SPDX-License-Identifier: MIT-0
+SPDX-License-Identifier: MIT-0
diff --git a/apigw-rest-stepfunctions-express-sync-bedrock-sam/apigw-rest-stepfunctions-express-sync-bedrock-sam.json b/apigw-rest-stepfunctions-express-sync-bedrock-sam/apigw-rest-stepfunctions-express-sync-bedrock-sam.json
@@ -1,17 +1,17 @@
 {
-    "title": "Prompt Chaining with Amazon API Gateway and AWS Step Functions",
-    "description": "Prompt Chaining no-code example with Amazon API Gateway HTTP API, AWS Step Functions and Amazon Bedrock.",
+    "title": "Prompt Chaining with Amazon API Gateway and AWS Step Functions.",
+    "description": "Prompt Chaining no-code example with Amazon API Gateway REST API, AWS Step Functions and Amazon Bedrock.",
     "language": "",
     "level": "200",
     "framework": "AWS SAM",
     "introBox": {
         "headline": "How it works",
         "text": [
             "The provided example demonstrates the deployment of a serverless application using the AWS Serverless Application Model (SAM) template.",
-            "This serverless application includes an Amazon API Gateway HTTP API endpoint that is connected to an AWS Step Functions state machine.",
+            "This serverless application includes an Amazon API Gateway REST API endpoint that is connected to an AWS Step Functions state machine.",
             "The state machine is designed to invoke an Express workflow synchronously.",
             "It utilizes AWS Step Functions' intrinsic functions to chain two prompts, which are then used to invoke the Amazon Bedrock language model.",
-            "The output from the state machine execution is returned to the client within 29 seconds through the HTTP API.",
+            "The output from the state machine execution is returned to the client within 29 seconds through the REST API.",
             "This no-code example showcases how the results from the first prompt can be used to provide context for the second prompt, allowing the language model to deliver a highly-curated response.",
             "By chaining these prompts, the system can leverage the capabilities of the language model to generate more meaningful and contextual outputs."
         ]
@@ -73,7 +73,7 @@
             "x": 20,
             "y": 50,
             "service": "apigw",
-            "label": "API Gateway HTTP API"
+            "label": "API Gateway REST API"
         },
         "icon2": {
             "x": 50,
@@ -90,12 +90,12 @@
         "line1": {
             "from": "icon1",
             "to": "icon2",
-            "label": ""
+            "label": "Request ANY {proxy+}"
         },
         "line2": {
             "from": "icon2",
             "to": "icon3",
-            "label": ""
+            "label": "Invoke"
         }
     }
 }
diff --git a/apigw-rest-stepfunctions-express-sync-bedrock-sam/example-pattern.json b/apigw-rest-stepfunctions-express-sync-bedrock-sam/example-pattern.json
@@ -1,17 +1,17 @@
 {
   "title": "Prompt Chaining with Amazon API Gateway and AWS Step Functions.",
-  "description": "Prompt Chaining no-code example with Amazon API Gateway HTTP API, AWS Step Functions and Amazon Bedrock.",
+  "description": "Prompt Chaining no-code example with Amazon API Gateway REST API, AWS Step Functions and Amazon Bedrock.",
   "language": "",
   "level": "200",
   "framework": "AWS SAM",
   "introBox": {
     "headline": "How it works",
     "text": [
       "The provided example demonstrates the deployment of a serverless application using the AWS Serverless Application Model (SAM) template.",
-      "This serverless application includes an Amazon API Gateway HTTP API endpoint that is connected to an AWS Step Functions state machine.",
+      "This serverless application includes an Amazon API Gateway REST API endpoint that is connected to an AWS Step Functions state machine.",
       "The state machine is designed to invoke an Express workflow synchronously.",
       "It utilizes AWS Step Functions' intrinsic functions to chain two prompts, which are then used to invoke the Amazon Bedrock language model.",
-      "The output from the state machine execution is returned to the client within 29 seconds through the HTTP API.", 
+      "The output from the state machine execution is returned to the client within 29 seconds through the REST API.",
       "This no-code example showcases how the results from the first prompt can be used to provide context for the second prompt, allowing the language model to deliver a highly-curated response.", 
       "By chaining these prompts, the system can leverage the capabilities of the language model to generate more meaningful and contextual outputs."
     ]
diff --git a/apigw-rest-stepfunctions-express-sync-bedrock-sam/statemachine/stateMachine.asl.json b/apigw-rest-stepfunctions-express-sync-bedrock-sam/statemachine/stateMachine.asl.json
@@ -6,42 +6,76 @@
       "Type": "Task",
       "Resource": "arn:aws:states:::bedrock:invokeModel",
       "Parameters": {
-        "ModelId": "cohere.command-text-v14",
+        "ModelId": "global.amazon.nova-2-lite-v1:0",
         "Body": {
-          "prompt.$": "$.prompt_one",
-          "max_tokens": 2000
-        },
-        "ContentType": "application/json",
-        "Accept": "*/*"
+          "messages": [
+            {
+              "role": "user",
+              "content": [
+                {
+                  "text.$": "$.prompt_one"
+                }
+              ]
+            }
+          ],
+          "inferenceConfig": {
+            "maxTokens": 2000
+          }
+        }
       },
       "Next": "Add first result to conversation history",
       "ResultPath": "$.result_one",
       "ResultSelector": {
-        "result_one.$": "$.Body.generations[0].text"
+        "result_one.$": "$.Body.output.message.content[0].text"
       }
     },
     "Add first result to conversation history": {
       "Type": "Pass",
       "Next": "Invoke model with second prompt",
       "Parameters": {
-        "convo_one.$": "States.Format('{}',$.result_one.result_one)"
+        "messages": [
+          {
+            "role": "user",
+            "content": [
+              {
+                "text.$": "$.prompt_one"
+              }
+            ]
+          },
+          {
+            "role": "assistant",
+            "content": [
+              {
+                "text.$": "$.result_one.result_one"
+              }
+            ]
+          },
+          {
+            "role": "user",
+            "content": [
+              {
+                "text": "Now write a short story based on the above."
+              }
+            ]
+          }
+        ]
       },
-      "ResultPath": "$.convo_one"
+      "ResultPath": "$.conversation"
     },
     "Invoke model with second prompt": {
       "Type": "Task",
       "Resource": "arn:aws:states:::bedrock:invokeModel",
       "Parameters": {
-        "ModelId": "cohere.command-text-v14",
+        "ModelId": "global.amazon.nova-2-lite-v1:0",
         "Body": {
-          "prompt.$": "States.Format('{}\n{}','Human: Now write a short story based on the following. Assistant:', $.convo_one.convo_one)",
-          "max_tokens": 1000
-        },
-        "ContentType": "application/json",
-        "Accept": "*/*"
+          "messages.$": "$.conversation.messages",
+          "inferenceConfig": {
+            "maxTokens": 2000
+          }
+        }
       },
       "ResultSelector": {
-        "result_two.$": "$.Body.generations[0].text"
+        "result_two.$": "$.Body.output.message.content[0].text"
       },
       "ResultPath": "$.result_two",
       "End": true
diff --git a/apigw-rest-stepfunctions-express-sync-bedrock-sam/template.yml b/apigw-rest-stepfunctions-express-sync-bedrock-sam/template.yml
@@ -1,6 +1,6 @@
 AWSTemplateFormatVersion: '2010-09-09'
 Transform: AWS::Serverless-2016-10-31
-Description: Prompt Chaining with Amazon API Gateway HTTP API, AWS StepFunctions and Amazon Bedrock.
+Description: Prompt Chaining with Amazon API Gateway REST API, AWS StepFunctions and Amazon Bedrock.
 
 Resources:
 
@@ -14,7 +14,7 @@ Resources:
       #Name: APIGWStepFunctionExpressSync
       DefinitionUri: statemachine/stateMachine.asl.json
       DefinitionSubstitutions:
-        ModelId: !Sub arn:aws:bedrock:${AWS::Region}::foundation-model/cohere.command-text-v14
+        ModelId: !Sub arn:aws:bedrock:${AWS::Region}:${AWS::AccountId}:inference-profile/global.amazon.nova-2-lite-v1:0
       Role:
         Fn::GetAtt: [ StatesMachineExecutionRole, Arn ]
       Type: EXPRESS
@@ -32,7 +32,7 @@ Resources:
     Type: AWS::Logs::LogGroup
     Properties:
       LogGroupName: !Join [ "/", [ "stepfunctions", StateMachineExpressSync]]
-       
+
 ##########################################################################
 #   REST API                                                             #
 ##########################################################################
@@ -73,7 +73,6 @@ Resources:
               Action: "states:StartSyncExecution"
               Resource: !GetAtt StateMachineExpressSync.Arn
 
- 
   StatesMachineExecutionRole:
     Type: "AWS::IAM::Role"
     Properties:
@@ -103,15 +102,17 @@ Resources:
                   - "logs:PutResourcePolicy"
                   - "logs:DescribeResourcePolicies"
                   - "logs:DescribeLogGroups"
-                Resource: "*"         
+                Resource: "*"
         - PolicyName: BedrockAccess
           PolicyDocument:
             Version: "2012-10-17"
             Statement:
               - Effect: Allow
                 Action:
                   - "bedrock:InvokeModel"
-                Resource: !Sub arn:aws:bedrock:${AWS::Region}::foundation-model/cohere.command-text-v14                           
+                Resource:
+                  - !Sub arn:aws:bedrock:${AWS::Region}:${AWS::AccountId}:inference-profile/global.amazon.nova-2-lite-v1:0
+                  - arn:aws:bedrock:*::foundation-model/amazon.nova-2-lite-v1:0
 
 ##########################################################################
 #   Outputs                                                              #
@@ -120,5 +121,3 @@ Outputs:
   PromptChainApi:
     Description: "Sync WF API endpoint"
     Value: !Sub "https://${RestApiforSyncWF}.execute-api.${AWS::Region}.amazonaws.com/dev"
-
-
