feat: add TokenLab provider

[hedging8563]

Summary

• add TokenLab as a cloud LLM provider using the OpenAI-compatible chat completions API
• load the live TokenLab model catalog from /v1/models and fall back to a curated set of current flagship models
• document TokenLab alongside the existing providers, including its native Responses, Anthropic Messages, and Gemini endpoints

Validation

• pnpm install --frozen-lockfile
• pnpm run typecheck
• pnpm run lint
• pnpm run build
• live TokenLab /v1/models readback: 353 total models, 140 chat models, 0 curated models missing