feat: add /v1/models fallback for LiteLLM when /v1/model/info is inaccessible

[roomote-v0]

Related GitHub Issue

Closes: #11898

Description

This PR attempts to address Issue #11898. Feedback and guidance are welcome.

When /v1/model/info is inaccessible (e.g. 403 Forbidden due to API gateways or internal policies), the getLiteLLMModels function now automatically falls back to /v1/models (the OpenAI-compatible endpoint) to retrieve the model list.

Key implementation details:

• Extracted URL building and header construction into reusable helpers (buildHeaders, buildUrl)
• Extracted response parsing into two exported functions: parseModelInfoResponse (for /v1/model/info) and parseModelsListResponse (for /v1/models)
• The main getLiteLLMModels function now tries /v1/model/info first. If it fails for any reason (HTTP error, network error, unexpected format), it catches the error, logs it, and retries with /v1/models
• The /v1/models fallback uses sensible defaults (8192 max tokens, 200k context window) since that endpoint provides less metadata
• If both endpoints fail, a descriptive error is thrown mentioning both endpoints failed
• Fully backward compatible: users whose /v1/model/info works will see no behavior change

Test Procedure

• All 30 existing and new tests pass via cd src && npx vitest run api/providers/fetchers/__tests__/litellm.spec.ts
• New test cases cover:
  • Fallback to /v1/models when /v1/model/info returns 403
  • Fallback to /v1/models when /v1/model/info returns 500
  • Fallback to /v1/models when /v1/model/info returns unexpected format
  • Both endpoints failing (HTTP errors, network errors, generic errors)
  • URL structure preserved in fallback requests
  • Invalid model IDs skipped in fallback response
  • Unit tests for parseModelInfoResponse and parseModelsListResponse parsers

Pre-Submission Checklist

• Issue Linked: This PR is linked to an approved GitHub Issue (see "Related GitHub Issue" above).
• Scope: My changes are focused on the linked issue (one major feature/fix per PR).
• Self-Review: I have performed a thorough self-review of my code.
• Testing: New and/or updated tests have been added to cover my changes.
• Documentation Impact: No documentation updates are required for this change.
• Contribution Guidelines: I have read and agree to the Contributor Guidelines.

Documentation Updates

• No documentation updates are required.

Additional Notes

This change helps enterprise/managed LiteLLM users where /v1/model/info may be blocked by API gateways, VPNs, or internal policies, while /v1/models remains accessible.

Interactively review PR in Roo Code Cloud