TokenLab OpenAI-Compatible Migration

Use this skill when a user wants to move an existing OpenAI-compatible integration, OpenRouter setup, LiteLLM config, LangChain/LlamaIndex app, or custom SDK wrapper to TokenLab.

What this skill should deliver

• A minimal diff or config change that preserves the existing user path.
• The correct TokenLab base URL and API key environment variable.
• A model migration note with live discovery commands.
• A native endpoint note when the old code actually needs Responses, Anthropic Messages, Gemini, media, audio, embeddings, rerank, or translation behavior.
• A rollback note that restores the previous provider config.

Default migration rules

• OpenAI-compatible SDKs:
  • base URL: https://api.tokenlab.sh/v1
  • auth: Authorization: Bearer $TOKENLAB_API_KEY
  • chat path: /v1/chat/completions
• Anthropic-native clients:
  • base URL: https://api.tokenlab.sh
  • messages path: /v1/messages
• Gemini-native clients:
  • base URL: https://api.tokenlab.sh
  • generateContent path: /v1beta/models/{model}:generateContent
• Live discovery:
  • GET https://api.tokenlab.sh/v1/models
  • GET https://api.tokenlab.sh/v1/models?recommended_for=<scene>
  • GET https://api.tokenlab.sh/v1/models/:model

Preferred approach

1. Locate the current provider config, environment variables, and model IDs.
2. Make the smallest possible change first:
   • replace base URL
   • replace API key variable
   • keep SDK/client shape intact
3. Add model discovery only where the app currently hardcodes provider-specific models.
4. Keep native semantics when the source integration uses Responses, Anthropic Messages, Gemini content parts, tools, media, audio, embeddings, rerank, or translation.
5. Provide a smoke test command for the migrated path.

Output format

• One short explanation of the migration surface.
• A focused diff, config block, or runnable replacement snippet.
• One smoke test command.
• One model discovery command.
• One rollback line naming the previous base URL/env vars.

Avoid

• Do not rewrite the user's SDK architecture if a base URL change is enough.
• Do not flatten native Anthropic or Gemini payloads into generic chat unless the user explicitly accepts that behavior change.
• Do not silently change models when the user's existing model has no obvious TokenLab equivalent.
• Do not embed API keys in source files or prompts.
• Do not claim feature parity without checking the model details endpoint.

Edge Cases

• If an app has separate base URLs for chat, embeddings, and media, migrate one family at a time.
• If the existing app uses OpenRouter model slugs, map to TokenLab public model IDs only after catalog lookup.
• If the user's previous provider had provider-specific headers, keep only headers that TokenLab documents or that the user's app still needs.