agent-craft

Generate Claude Code agent definitions with roles, tools, and system prompts.

Updated Apr 8, 2026
One-click install
npx skills add https://github.com/Gami19/rd-strands-agents --skill agent-craft
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-craft
Source: https://github.com/Gami19/rd-strands-agents/tree/main/strands-py/backend/agent/skills/pr/agent-craft
Command: npx skills add https://github.com/Gami19/rd-strands-agents --skill agent-craft

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the end-to-end design and generation of customized Claude Code agents, turning user requirements into deployable agent definitions.

Core Features & Use Cases

  • Generate agent roles, tool constraints, model choices, and system prompts from user input
  • Output ready-to-deploy .md files under .claude/agents/ with design rationales and test plans
  • Suitable for multi-agent workflows in development, QA, and operations

Quick Start

Provide a complete agent definition by specifying the task, scope, and outputs and generate the .claude/agents/<name>.md file.

Frequently Asked Questions about agent-craft

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a custom Claude Code agent for a multi-agent workflow?

To generate a custom Claude Code agent, you provide the specific task requirements, scope, and desired outputs. The Skill then designs agent roles, tool constraints, and model choices, outputting a ready-to-deploy .md agent file directly into your .claude/agents/ directory.

What is a .claude/agents/<name>.md file used for?

A .claude/agents/<name>.md file is a ready-to-deploy agent definition used to configure customized Claude Code agents. It contains the system prompt, design rationale, tool constraints, and test plans required for the agent to function within multi-agent workflows.

Can I specify tool constraints and model choices when designing Claude Code agents?

Yes, you can specify tool constraints and model choices when designing Claude Code agents. The generation process takes your user requirements and translates them into specific tool restrictions and model configurations within the outputted system prompt.

Does the generated Claude Code agent file include test plans and validation guidance?

Yes, the generated Claude Code agent file includes test plans and validation guidance. Alongside the deployable agent definition, the output documents the design rationale and provides specific instructions for integrating and validating the agent in your workflow.

What's the best way to automate QA pipeline agent creation for Claude Code?

The best way to automate QA pipeline agent creation is to input your testing requirements and scope into this Skill. It automatically generates the tailored agent roles and system prompts needed, outputting a validated .md file ready for immediate integration into your QA pipelines.