About
This skill provides a specialized framework for integrating the Claudish CLI with Claude Code, allowing developers to leverage alternative models like Grok, GPT-5, and Gemini while strictly maintaining context window hygiene. It includes a mandatory decision tree for agent delegation, specialized file-based instruction patterns, and a curated selection matrix to ensure that external model calls happen within isolated sub-agents. By following these patterns, users can prevent token pollution in their main conversation while harnessing the specialized capabilities of various state-of-the-art language models.
