dyad:add-models

Add AI model entries to the language model constants file with researched specifications.

21.4k|2.6k|Updated Apr 11, 2025
One-click install
npx skills add https://github.com/dyad-sh/dyad --skill dyad-add-models
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dyad:add-models
Source: https://github.com/dyad-sh/dyad/tree/main/.claude/skills/add-models
Command: npx skills add https://github.com/dyad-sh/dyad --skill dyad-add-models

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Adding new AI models to Dyad's language model constants file requires manually researching each model's API name, context window, token limits, and pricing tier across multiple provider documentation sites, which is error-prone and time-consuming.

Core Features & Use Cases

  • Multi-Provider Model Registration: Add models for Anthropic, OpenAI, Google, xAI, OpenRouter, Azure, Bedrock, and Vertex into src/ipc/shared/language_model_constants.ts following each provider's naming and configuration conventions.
  • Automated Spec Research: Uses web search to look up official documentation for API model names, max output tokens, context windows, default temperatures, and cost tiers.
  • Convention Enforcement: Applies provider-specific patterns such as ARN formats for Bedrock, prefixed names for OpenRouter, and correct temperature defaults, then verifies changes compile with npm run ts.
  • Use Case: When a new model like a Gemini or Claude release comes out, run this Skill with the model names to insert properly formatted entries, update related arrays like TURBO_MODELS, and create named constant exports where needed.

Quick Start

Add the models "gemini 3.1 pro, glm 5" to the language model constants file using this Skill.

Frequently Asked Questions about dyad:add-models

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add a new AI model to Dyad's supported models list?

Provide a comma-separated list of model names as arguments, and the Skill researches each model's specs from official docs, then inserts properly formatted entries into the appropriate provider arrays in language_model_constants.ts.

How are model pricing tiers determined when adding models?

Pricing tiers use a dollar-sign scale from 0 to 6 based on per-million-token input pricing, ranging from free (0) through cheap tiers under $2/M up to premium tiers above $30/M, relative to other models in the same provider.

Which AI providers does the model constants file support?

The constants file supports Anthropic, OpenAI, Google, xAI, OpenRouter, Azure, Bedrock, and Vertex. Each provider has distinct conventions, such as ARN-format names for Bedrock and provider-prefixed names for OpenRouter.

Can one model be added to multiple providers at once?

Yes, a model like a new Anthropic release can belong to both the anthropic and bedrock providers. The Skill adds it to the primary provider and asks whether you also want it registered under secondary providers.

How are changes to the model constants file verified?

After inserting new model entries, the Skill runs npm run ts to type-check the codebase and fixes any compilation errors. It also checks whether related arrays like TURBO_MODELS or named constant exports need updating.