fix(core): handle MiMo tool-result media#4281
Open
DragonnZhang wants to merge 2 commits into
Open
Conversation
Contributor
📋 Review SummaryThis PR introduces a new MiMo OpenAI-compatible provider that addresses two specific issues with the MiMo API: (1) preserving 🔍 General Feedback
🎯 Specific Feedback🟡 High Priority
function ensureReasoningContentOnToolCalls(
message: OpenAI.Chat.ChatCompletionMessageParam,
): OpenAI.Chat.ChatCompletionMessageParam {
// ... logic ...
return {
...assistant,
reasoning_content: '',
} as OpenAI.Chat.ChatCompletionMessageParam; // This cast drops reasoning_content from the type
}Suggested fix: Either return 🟢 Medium Priority
const providerOverrides =
this.config.provider.getRequestContextOverrides?.() ?? {};
// Consider: const providerOverrides: OpenAIRequestContextOverrides =
// this.config.provider.getRequestContextOverrides?.() ?? {};🔵 Low Priority
getRequestContextOverrides(): { splitToolMedia?: boolean } {
// Respect explicit user configuration; default to true for MiMo compatibility
return {
splitToolMedia: this.contentGeneratorConfig.splitToolMedia ?? true,
};
}
✅ Highlights
|
Contributor
Code Coverage Summary
CLI Package - Full Text ReportCore Package - Full Text ReportFor detailed HTML reports, please see the 'coverage-reports-22.x-ubuntu-latest' artifact from the main CI run. |
wenshao
reviewed
May 18, 2026
wenshao
reviewed
May 18, 2026
wenshao
requested changes
May 18, 2026
This was referenced May 20, 2026
wenshao
reviewed
May 20, 2026
| createCliConfig(), | ||
| ); | ||
|
|
||
| expect(provider).toBeInstanceOf(MiMoOpenAICompatibleProvider); |
Collaborator
There was a problem hiding this comment.
[Suggestion] Missing test for determineProvider ordering when both MiMo model name and OpenRouter hostname match.
A user configuring baseUrl: 'https://openrouter.ai/api/v1' + model: 'mimo-v2.5-pro' currently gets the MiMo provider (checked before OpenRouter in determineProvider). No test asserts this priority. If someone later reorders the checks, users routing MiMo models through OpenRouter would silently get the wrong provider — losing reasoning_content injection and splitToolMedia: true.
Consider adding:
it('is selected over OpenRouter when both hostname and model match', () => {
const provider = determineProvider(
createProviderConfig({
baseUrl: 'https://openrouter.ai/v1',
model: 'mimo-v2.5-pro',
}),
createCliConfig(),
);
expect(provider).toBeInstanceOf(MiMoOpenAICompatibleProvider);
});— qwen-latest-series-invite-beta-v34 via Qwen Code /review
wenshao
approved these changes
May 20, 2026
Collaborator
wenshao
left a comment
wenshao
left a comment
There was a problem hiding this comment.
No issues found. LGTM! ✅ — qwen-latest-series-invite-beta-v34 via Qwen Code /review
Collaborator
Maintainer test report — PR #4281Built and validated locally in a dedicated Environment
Results
Pre-existing failure (NOT caused by this PR)Reproduced on a clean worktree of Behavioral checks against the diff
Risk assessmentLow risk for merge.
Not covered locally
Reproducegit fetch origin pull/4281/head:pr-4281 && git checkout pr-4281
npm install
cd packages/core && npx vitest run \
src/core/openaiContentGenerator/provider/mimo.test.ts \
src/core/openaiContentGenerator/pipeline.test.ts
cd ../.. && npm run typecheckRecommendation: safe to merge once at least one reviewer signs off on the |
5 tasks
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
3 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.
Summary
role: \"tool\"messages.400 Param Incorrectwhen tool-call history omitsreasoning_content, and strict chat-completions servers can also reject image/video/file content embedded directly in tool result messages.splitToolMediadefault for MiMo, and the request-context override precedence.Validation
1f1a7c9d-f054-46fd-8458-1717cd611968, where the user asked Qwen Code to read any repository image andread_filereturned a PNG as inline media.reasoning_contentfor assistant tool-call turns.packages/coretypecheck passed.read_fileand verify the follow-up request no longer fails with400 Param Incorrect.read_filereturninginlineDataforicon-orange.png, followed immediately byqwen-code.api_error status=400 model=mimo-v2.5 msg=400 Param Incorrect.Scope / Risk
splitToolMedia: falseto opt out.Testing Matrix
Testing matrix notes:
npx vitestandpackages/corenpm run typecheck.npm run typecheckwas attempted but is currently blocked by unrelated existingpackages/cli/src/serve/workspaceAgents.tsandworkspaceMemory.tstype errors.Linked Issues / Bugs
Fixes #4223