fix(providers): update SenseNova Token Plan support - #6330

Summary

• Update the built-in SenseNova provider to the Token Plan OpenAI-compatible endpoint: https://token.sensenova.cn/v1/chat/completions.
• Replace the legacy SenseCore/SenseChat catalog with the smoke-tested Token Plan chat models: sensenova-6.7-flash-lite, deepseek-v4-flash, and glm-5.2.
• Encode SenseNova's observed 65,536 max output token ceiling in both registry model capabilities and provider-level request clamping, so high-output combo paths do not forward 96k max_tokens to SenseNova.
• Lock key validation to the working OpenAI-compatible /v1/models endpoint derived from the corrected Token Plan base URL.

Related Issues

• Related to provider key validation and SenseNova Token Plan runtime compatibility.

Validation

• npm run lint
• npm run test:unit
• npm run test:coverage
• Coverage is still >= 60% for statements, lines, functions, and branches
• SonarQube PR analysis is green or any remaining issues are explicitly documented below

Targeted validation run locally:

• DISABLE_SQLITE_AUTO_BACKUP=true node --max-old-space-size=8192 --import tsx --import ./open-sse/utils/setupPolyfill.ts --import ./tests/_setup/isolateDataDir.ts --test --test-force-exit tests/unit/model-capabilities-registry.test.ts tests/unit/sensenova-provider.test.ts
• npm run typecheck:core
• npm run check:provider-assets
• pre-commit checks on commit/push: prettier, eslint, docs sync, T11 any-budget, tracked-artifacts

Not run locally:

• npm run check:provider-consistency requires bun; this local environment does not have bun installed.

Tests Added Or Updated

• Added tests/unit/sensenova-provider.test.ts covering:
  • Token Plan endpoint selection
  • exposed SenseNova Token Plan chat models
  • model capability metadata and 65,536 max output token cap
  • resolved max output token clamp from 96,000 to 65,536
  • provider-level request cap
  • key validation through https://token.sensenova.cn/v1/models

Coverage Notes

• Production changes are in open-sse/config/providers/registry/sensenova/index.ts and open-sse/config/constants.ts.
• New unit tests directly import the registry entry, capability resolver, provider cap constant, and provider key validation function.

Reviewer Notes

• No SenseNova test key is included in code, tests, logs, or fixtures.
• sensenova-u1-fast is intentionally not exposed as a chat model because it is currently advertised by /v1/models but returns chat-completions 404 for the tested Token Plan account.
• The corrected /v1/models route exists: unauthenticated probing returns 401 Authorization Not Found, not a provider-validation unsupported response.