model-registry-maintainer

The 'model-registry-maintainer' skill is designed to help users maintain and update the MassGen model registry, which includes both the model capabilities and pricing/token management. The skill is essential for ensuring the registry is up-to-date with the latest model releases, pricing changes, context window updates, and other key model information. It serves as a critical tool for keeping the backend and capabilities registry current with new models or changes to existing ones from different providers such as OpenAI, Anthropic, Google, and xAI.

This skill is primarily composed of two files: 'capabilities.py', which tracks model features, release dates, and capabilities, and 'token_manager.py', which manages model pricing and context window limits. Key actions include adding new models, updating pricing and context windows, and ensuring that any changes from providers are reflected in the registry. It also provides a clear process for integrating new models by gathering relevant information like release dates, pricing, and capabilities, and checking if the model is listed in the LiteLLM database to avoid redundant updates.