model-detection

Model detection identifies the active LLM (Claude Sonnet, Claude Haiku, Claude Opus, or GLM-4.6) through system context analysis, performance pattern recognition, and capability assessment. This enables the agent to adapt execution strategies, skill loading, and communication style for optimal compatibility and performance.

Yes. Model-specific configurations tailor reasoning style, context management, and error handling to each LLM's strengths. For example, Claude Haiku 4.5 triggers concise communication and efficient skill loading, while other models receive alternative configurations optimized for their performance patterns.

Use capability testing functions to verify model characteristics including nuanced reasoning and structured execution. The Skill assesses these capabilities upon initialization and automatically configures adaptive execution—progressive or complete skill loading, natural or structured communication—ensuring reliable operation across Claude Sonnet 4.5, Claude Haiku 4.5, Claude Opus 4.1, and GLM-4.6.

System context analysis examines environmental indicators and model-specific behavioral patterns. Combined with performance pattern recognition, it accurately determines model identity without manual configuration, allowing autonomous agents to apply the correct optimization strategies immediately upon startup.

Model detection is most valuable for agents deployed across multiple LLMs or environments where the active model may vary. If your use case is single-model only, detection provides minimal benefit, though it enables future portability and reduces reconfiguration overhead if you later migrate to a different LLM.

Reliable error handling is enforced as a functional requirement. If detection fails, the agent falls back to safe defaults or requests explicit model specification. The Skill includes capability testing to verify detection accuracy and trigger recalibration if model characteristics don't match expectations.