feat: add MiniMax-M2.7 model pricing support

[octo-patch]

Summary

• Add MiniMax-M2.7 and MiniMax-M2.7-highspeed to FALLBACK_PRICING in src/models.ts with accurate per-token costs
• Add human-readable display names (MiniMax M2.7, MiniMax M2.7 Highspeed) in getShortModelName
• Add unit tests for MiniMax model pricing and display name resolution in tests/minimax.test.ts

This enables CodeBurn to accurately report costs when users run MiniMax models through providers like OpenCode that support configurable LLM backends.

Pricing (from MiniMax API pricing):

Model	Input	Output	Cache Write	Cache Read
MiniMax-M2.7	$0.3/M	$1.2/M	$0.375/M	$0.06/M
MiniMax-M2.7-highspeed	$0.6/M	$2.4/M	$0.375/M	$0.06/M

Test plan

• npm test -- --run tests/minimax.test.ts - all 6 tests pass
• Pricing for MiniMax-M2.7 returns correct per-token costs from FALLBACK_PRICING
• Pricing for MiniMax-M2.7-highspeed returns higher costs than base model
• getShortModelName returns MiniMax M2.7 / MiniMax M2.7 Highspeed
• Model IDs with date suffixes are handled correctly

References

• Chat API: https://platform.minimax.io/docs/api-reference/text-openai-api
• Pricing: https://platform.minimax.io/docs/guides/pricing-paygo

[iamtoruk]

Thanks for the PR. We've already got MiniMax support on our pipeline and will ship the pricing plus display name additions in the next release. Closing this one. Appreciate you putting the tests together.