fix: correct Mistral codestral-latest context window to 131k

[roomote]

Summary

This PR attempts to address Issue #9344 by correcting the context window limit for Mistral's codestral-latest model.

Problem

The codestral-latest model was incorrectly configured with a context window of 256,000 tokens when the official Mistral documentation specifies it should be 131,000 tokens (128k in practice).

Solution

Updated the model configuration in packages/types/src/providers/mistral.ts to change:

• maxTokens: from 256,000 to 131,000
• contextWindow: from 256,000 to 131,000

Testing

• ✅ All existing tests pass
• ✅ Ran Mistral provider tests: npm test api/providers/__tests__/mistral.spec.ts
• ✅ Ran types package tests: npm test in packages/types
• ✅ Linting and type checking pass

Fixes #9344

Feedback and guidance are welcome!

---

Important

Corrects maxTokens and contextWindow for codestral-latest model in mistral.ts to 131,000.

• Behavior:
  • Corrects maxTokens and contextWindow for codestral-latest model in mistral.ts from 256,000 to 131,000.
• Testing:
  • All existing tests pass.
  • Ran Mistral provider tests: npm test api/providers/__tests__/mistral.spec.ts.
  • Ran types package tests: npm test in packages/types.
  • Linting and type checking pass.

^{This description was created by for 7cac64d. You can customize this summary. It will automatically update as commits are pushed.}

[roomote]

Rooviewer   See task on Roo Cloud

Review complete. No issues found. The PR correctly updates the codestral-latest model's context window from 256,000 to 131,000 tokens, aligning with Mistral's official documentation.

_{Mention @roomote in a comment to request specific changes to this pull request or fix all unresolved issues.}