Description

#1078 mentioned that context overflows were not handled, but I wasn't able to reproduce using the code changes in it. However, in testing (using @DEA's suggested test) I was able to reproduce and consistently got a "prompt is too long:" error. This change addresses that error.

I also added a test case using the expected error and validated that it works. If we swap out ConversationManagers, the agent loop does automatically trim the context (not propagating the error) as expected.

B​est practices to handle prompts that are too long for the LLM API (eg., Anthropic, OpenAi)? - DEV Community also indicates that this is an error that is not unique to us.

Related Issues

N/A

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.