feat(llm): list available tools in unknown-function error output#6085
Open
ShayneP wants to merge 1 commit into
Open
feat(llm): list available tools in unknown-function error output#6085ShayneP wants to merge 1 commit into
ShayneP wants to merge 1 commit into
Conversation
When a model hallucinates a tool name, the error fed back to it only said "Unknown function: <name>", giving it nothing to correct with. Include the available tool names in the error output so the model can retry with a valid tool on the next turn.
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
theomonnom
approved these changes
Jun 12, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
When a model hallucinates a tool name, the error fed back to it is just
Unknown function: <name>, which gives it nothing to self-correct with — smaller models in particular tend to retry the same invalid name.Include the available tool names in the error output (both in
execute_function_calland the voice generation path) so the model can pick a valid tool on the next attempt.