Description

In some cases, LLM decides to respond nothing given the tool has done the job, which causes empty assistant content and received validation exception from bedrock models. Therefore, appending "blank text" when the whole content is empty.

Possible Solutions

Option 1: Mutate the message when receiving the response
Option 2 (chosen): update the message before calling the model

The benefit of option 1 is to make sure that no message with empty content is stored/used. However, we will not be able to see the actual response from the model. Therefore we decided to go for option 2 in order to keep the conversatin history as origin as possible.

Related Issues

#514
strands-agents/tools#190

Documentation PR

N/A

Type of Change

Bug fix

Testing

How have you tested the change? Verify that the changes do not break functionality or introduce warnings in consuming repositories: agents-docs, agents-tools, agents-cli

• I ran hatch run prepare

Checklist

• I have read the CONTRIBUTING document
• I have added any necessary tests that prove my fix is effective or my feature works
• I have updated the documentation accordingly
• I have added an appropriate example to the documentation to outline the feature, or no new docs are needed
• My changes generate no new warnings
• Any dependent changes have been merged and published

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.