Repository navigation
Conversation
|
|
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Use the documented
1.0USD default when a Claude Code model'smax_budget_usdis omitted ornull.Real
LLMModelConfigobjects have this attribute set toNoneby default. The provider'sgetattr(..., 1.0)fallback therefore does not apply, and the generated command contains--max-budget-usd None. Existing backend tests supplied a numeric budget through a mock config, hiding the problem.This change handles only
None; explicit values and per-call overrides keep their existing behavior. Five regression tests exercise the real config loader and provider factory with the subprocess mocked.Validation
git diff --check: passed.The full unit suite ran with test-only guards preventing external network requests and real model CLI subprocesses; its local fake HTTP server was allowed. No live Claude Code invocation or paid model call was tested. The model-backed integration suite was not run.