Guide AI agent architecture design with tool design and multi-agent coordination.
npx skills add https://github.com/fauzanazz/claude-skills --skill agent-design-fauzanazz
Please help me install this Agent Skill. Skill: agent-design Source: https://github.com/fauzanazz/claude-skills/tree/main/agent-design Command: npx skills add https://github.com/fauzanazz/claude-skills --skill agent-design-fauzanazz
This Skill offers practical guidance for designing and evaluating AI agent architectures, streamlining the process of creating robust and efficient agents.
To start designing an agent, use the skill 'agent-design' with the prompt "How do I design an AI agent for the task of X?".
High-intent search queries and answers about installing and using this skill.
To design an AI agent architecture, you need a framework for model-fit, tool design, and progressive disclosure. This Skill provides guidelines to analyze requirements and create robust agents optimized for the Claude API.
Progressive disclosure in AI agent design is a technique to manage context and tool complexity by revealing information only as needed. This Skill provides guidelines to implement it for optimizing agent behavior.
The best way to coordinate multiple AI agents is using structured strategies for collaboration. This Skill offers specific techniques for multi-agent coordination to ensure efficient task delegation and execution.
Yes, you can review existing agent architectures. This Skill provides a design review framework to analyze your current agent setup and offers techniques to improve tool design and model-fit.
While the guidelines focus on Claude API agents, the core principles of tool design, progressive disclosure, and multi-agent coordination are broadly applicable to evaluating and optimizing AI agent behavior across different platforms.
A limitation of progressive disclosure in tool design is the potential for context fragmentation if not managed carefully. This Skill provides guidelines to balance information availability with agent reasoning limits.
fableindigo-gifRemediate critical CVEs and implement secure-by-default patterns in claude-flow v3.
faionfaionGuide AI agent development with LangChain and LlamaIndex architectures.
AnveioIntegrate Claude models in TypeScript applications using the Anthropic SDK.
Tonybleything76Design multi-agent system architectures and generate tool schemas.
