What is the purpose of the model-routing skill?

It provides Claude with specific instructions on which model version (Sonnet 4.5 or Haiku 4.5) to use for different types of development tasks to ensure the best balance of reasoning depth and efficiency.

Does this skill affect all files in the repository?

Yes, these model routing guidelines apply to all tasks within the html-to-markdown project unless a specific task explicitly overrides them.

When does Claude use Sonnet 4.5 in this project?

Sonnet 4.5 is used for complex tasks including Rust core architecture, safety-critical code changes, cross-cutting refactoring, and algorithm optimization.

When is Haiku 4.5 preferred over Sonnet?

Haiku 4.5 is the preferred choice for writing tests, managing CLI tools, updating dependencies, and engineering bindings for languages like Python, TypeScript, and PHP.

Codex AI Model Routing

Name: Codex AI Model Routing
Author: kreuzberg-dev

bykreuzberg-dev

•

480

데이터 과학 및 ML

Optimizes development efficiency by routing complex Rust architecture tasks to Sonnet 4.5 and routine utility work to Haiku 4.5.

소개

This skill provides a strategic framework for using Claude within the html-to-markdown repository, ensuring that AI resources are utilized effectively. By directing high-complexity tasks like Rust core design and safety-critical refactoring to Sonnet 4.5, and assigning routine tasks such as test writing, CLI development, and language binding maintenance to Haiku 4.5, it maximizes code quality and developer throughput. This model-aware approach helps maintain the high performance and CommonMark compliance standards expected of the Kreuzberg document intelligence engine.

주요 기능

Standardized strategies for error handling and dependency management
Optimized guidance for Rust core architecture and algorithm design
Efficient handling of multi-language bindings (Python, TS, Ruby, PHP)
Streamlined CI/CD and build system integration workflows
Intelligent model selection based on task complexity and technical risk
480 GitHub stars

사용 사례

Refactoring safety-critical Rust core components using Sonnet 4.5
Implementing complex algorithm optimizations for CommonMark compliance
Generating unit tests and managing document fixtures with Haiku 4.5

소개

주요 기능

Standardized strategies for error handling and dependency management
Optimized guidance for Rust core architecture and algorithm design
Efficient handling of multi-language bindings (Python, TS, Ruby, PHP)
Streamlined CI/CD and build system integration workflows
Intelligent model selection based on task complexity and technical risk
480 GitHub stars

사용 사례

Refactoring safety-critical Rust core components using Sonnet 4.5
Implementing complex algorithm optimizations for CommonMark compliance
Generating unit tests and managing document fixtures with Haiku 4.5