adding-models

What problem does it solve?

This Skill guides you through adding support for new LLM models to Letta Code, including updating src/models.json, CI test matrices, and model validation. It reduces configuration errors and speeds up onboarding of new providers.

Core Features & Use Cases

• Model discovery: Identify valid model handles from common provider prefixes (e.g., openai/, anthropic/, google_ai/).
• Configuration updates: Add entries to src/models.json with fields like id, handle, label, description, and provider-specific updateArgs.
• CI & validation: Extend CI test matrices and validate model handles to prevent misconfigurations.
• Use Case: When you want to add support for a new model (e.g., a new Claude, GPT, or Gemini variant), this skill walks you through the exact steps.

Quick Start

Use the skill to add a new model by updating the models configuration and ensuring CI validation. Example commands:

• Update src/models.json with a new model entry
• Run validation to verify the handle and configuration
• Run CI tests to confirm compatibility