develop-agent

Develop Claude Code agent definitions from user requirements via guided interviews.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/christophevg/c3 --skill develop-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: develop-agent
Source: https://github.com/christophevg/c3/tree/main/skills/develop-agent
Command: npx skills add https://github.com/christophevg/c3 --skill develop-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It solves the problem of turning an idea for a Claude Code agent into a clear, testable, and production-ready agent definition with proper scope, tools, and guardrails.

Core Features & Use Cases

  • Step-by-step agent development workflow: Guides an end-to-end process from initial requirements interview through analysis, agent design, creation, and validation.
  • Scope and reliability controls: Enforces a narrow-scope “Three Tests for Narrow Scope” (trigger, action/output, and failure containment) to prevent overly broad or brittle agents.
  • Reusable agent blueprint: Produces an agent artifact using a five-section system-prompt framework and an agent frontmatter template that includes tools, output expectations, and error handling.
  • When to use: Use it when you want to create a new specialized agent, review an existing agent’s structure, or improve an agent’s scope/tooling and documentation.

Quick Start

Ask it to create an agent that can analyze API responses and produce a structured report with clearly defined inputs, outputs, and tool permissions.

Frequently Asked Questions about develop-agent

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

FAQPage Schema
How do I create a Claude Code agent from scratch?

To create a Claude Code agent, this Skill guides an end-to-end workflow from an initial requirements interview through functional analysis and design, ultimately producing a validated agent markdown artifact with YAML frontmatter.

What is the best way to prevent scope creep when developing specialized agents?

To prevent scope creep when developing specialized agents, the workflow enforces least-privilege tool selection and applies the Three Tests for Narrow Scope, validating trigger conditions, action outputs, and failure containment.

How does an agent development workflow formalize a bounded task?

An agent development workflow formalizes a bounded task by conducting a requirements interview, designing a five-section system-prompt framework, and generating an agent artifact with clearly defined inputs, outputs, and error handling.

Can I use this to review and improve an existing agent's structure and guardrails?

Yes, you can review an existing agent's structure and improve its clarity, scope, and guardrails by applying the scope validation process and enforcing least-privilege tool selection within the agent definition.

What format does the generated agent artifact use for Claude Code?

The generated agent artifact uses a markdown file format with YAML frontmatter, incorporating a five-section system-prompt framework that specifies tools, output expectations, and error handling.