model-detection

Overview

This skill provides universal model detection and capability assessment to optimize the Autonomous Agent Plugin across different LLM models (Claude Sonnet, Claude 4.5, GLM-4.6, etc.).

Model Detection Algorithm

Primary Detection Methods

1. System Context Analysis:

   // Check for model indicators in system context
   const modelIndicators = {
     'claude-sonnet-4.5': { pattern: /sonnet.*4\.5|4\.5.*sonnet/i, confidence: 0.9 },
     'claude-haiku-4.5': { pattern: /haiku.*4\.5|4\.5.*haiku/i, confidence: 0.9 },
     'claude-opus-4.1': { pattern: /opus.*4\.1|4\.1.*opus/i, confidence: 0.9 },
     'glm-4.6': { pattern: /glm|4\.6/i, confidence: 0.9 },
     'claude-haiku': { pattern: /haiku(?!\.*4\.5)/i, confidence: 0.8 }
   }