S3 is a simple AWS Lambda function that allows for integration between HttpApi (v2) and S3. This integration is currently not supported on HttpApi resulting in need to use HttpIntegration and having bucket on public-read which is not desired.
This function circumvents this limitation by standing between ApiGateway and S3 and proxying events.
Very simple and straight forward npx --package @sladg/s3-lambda cli deploy. Pass --help in the end for all options and configuration.
In case you want to clean it up, use npx --package @sladg/s3-lambda cli remove.
Command will output "s3LambdaArn" that you can use to make requests.
We are using pathParameters.proxy as key for S3, so you should set your HttpApi with /prefix/{proxy+} route, where prefix won't be passed to S3. There is currently no setting for using multiple S3 buckets or setting prefix for S3 key.