agent-artifex:design

Effective tool descriptions for LLM agent development require clear, evidence-based principles that help the model accurately interpret tool functionality. This Skill provides structured guidance and empirical best practices for writing descriptions that improve AI service reliability and actionability.

Parameter and schema design in AI agents should adhere to proven empirical principles to ensure robust system reliability. This Skill offers detailed, evidence-based guidance for defining parameters and constructing schemas that LLMs can easily interpret and process correctly.

Error messages for AI agent tools should be structured to allow the LLM to easily interpret and act upon the failure. This Skill provides evidence-based design principles for constructing error messages that enable agents to recover autonomously and maintain multi-turn conversation flow.

Multi-turn conversation architecture for LLMs is best designed using evidence-based principles that standardize response formats and tool set organization. This Skill delivers empirical guidance for structuring system prompts and managing complex agent interactions over multiple turns.

Building high-quality AI services relies heavily on empirical research data to ensure reliability and maintainability. This Skill supplies an evidence-based framework, integrating proven design principles for tool sets and prompt engineering, eliminating the need to gather research independently.