A serverless web and mobile backend uses Lambda and API Gateway. The backend Lambda deployment must be fully automated based on code pushed to the appropriate branch in CodeCommit. Requirements: separate pipelines for test and production, and automatic deploys only for test environments. What steps meet these requirements?

The correct option meets all requirements. Two CodePipeline pipelines (test and production) ensure separate deployment flows. A manual approval step in the production pipeline prevents automatic deploys to production, fulfilling the "automatic deploys only for test environments" requirement. Using a single CodeCommit repository with branches for each environment is a standard and efficient way to manage code for different stages. Configuring each pipeline to pull from its appropriate branch ensures the correct code version is deployed to each environment. CloudFormation is the recommended AWS service for deploying and managing serverless applications, including Lambda functions. Incorrect options: Creating separate CodeCommit repositories per environment is unnecessary and adds management overhead. The option involving CodeBuild projects doesn't fully describe a complete CI/CD pipeline and pushing Lambda code to S3 directly for deployment is less robust than using CloudFormation.