How does it handle prompt differences between models?

It provides specific snippets to address common Opus 4.5 issues like over-engineering, tool overtriggering, and 'thinking' sensitivity if the user reports these behaviors.

Does this skill support Haiku migration?

No, this migration assistant specifically excludes Haiku models and focuses on the Opus 4.5 transition.

Which models can I migrate from using this skill?

This skill supports migrating from Claude Sonnet 4.0, Sonnet 4.5, and Opus 4.1 to the newer Opus 4.5 model.

What is the 'effort' parameter mentioned in the migration?

Opus 4.5 supports an effort parameter; this skill helps set it to 'high' to ensure the model utilizes its maximum reasoning capabilities.

Which cloud platforms are covered in the model string updates?

It handles model strings for the Anthropic API (1P), AWS Bedrock, Google Vertex AI, and Azure AI Foundry.

Opus 4.5 Migration Assistant

Name: Opus 4.5 Migration Assistant
Author: Microck

byMicrock

•

API Development

Automates the migration of codebases and prompts from Claude Sonnet and Opus 4.1 to the latest Opus 4.5 model.

About

This skill facilitates a seamless transition to Claude Opus 4.5 by identifying and updating model strings across various platforms including AWS Bedrock, Google Vertex AI, and Azure AI Foundry. Beyond simple string replacement, it provides intelligent prompt adjustments to address behavioral differences unique to Opus 4.5, such as tool overtriggering, over-engineering tendencies, and code exploration hurdles. It ensures your integration utilizes the latest features like the 'high' effort parameter while maintaining stability by removing deprecated beta headers and refining thinking sensitivity instructions.

Key Features

Deprecated beta header removal and management (e.g., 1M context headers)
Intelligent prompt tuning to prevent tool overtriggering and over-engineering
Implementation of the 'high' effort parameter for enhanced reasoning
81 GitHub stars
Optimized instructions for frontend design quality and code exploration
Automated model string replacement for Anthropic, AWS, Google, and Azure platforms

Use Cases

Upgrading a production LLM application from Claude 3.5 Sonnet to Opus 4.5
Refining system prompts to handle the specific behavioral changes of the Opus 4.5 architecture
Standardizing model API calls across multi-cloud environments like AWS Bedrock and Vertex AI

About

Key Features

Deprecated beta header removal and management (e.g., 1M context headers)
Intelligent prompt tuning to prevent tool overtriggering and over-engineering
Implementation of the 'high' effort parameter for enhanced reasoning
81 GitHub stars
Optimized instructions for frontend design quality and code exploration
Automated model string replacement for Anthropic, AWS, Google, and Azure platforms

Use Cases

Upgrading a production LLM application from Claude 3.5 Sonnet to Opus 4.5
Refining system prompts to handle the specific behavioral changes of the Opus 4.5 architecture
Standardizing model API calls across multi-cloud environments like AWS Bedrock and Vertex AI