Build Agent

This is the agent that builds other agents. It walks you through an adaptive Q&A that only asks what's missing from your initial description, checks for overlap with existing agents to avoid duplication, then generates a complete markdown file following strict structural conventions: frontmatter fields in the right order, DO_NOT_TOUCH zones wrapping core instructions, MAY_EDIT sections for project config, and a self-critique checklist before showing you anything. It knows about tool scoping, memory modes, permission levels, and will update your orchestrator agent if you're building a core team member. The attention to YAML pitfalls and the "no em-dashes" rule suggest this emerged from real build pain.

GitHub stars: 231
last updated: Apr 7, 2026

View on GitHub

Install

$ npx -y skills add lukerenton/explore-claude-code --skill build-agent --agent claude-code

Installs into .claude/skills of the current project.

Files

You are an agent builder. You create well-structured, consistent Claude Code agent files that follow established standards. Your output is a complete, ready-to-deploy agent markdown file.

Process

1. Assess Input Completeness

Read what the user provided in $ARGUMENTS and the surrounding conversation context. Determine how much is already specified vs what needs clarification.

Categorize as:

Minimal (just a role name or vague idea): Full Q&A needed
Medium (clear role but missing configuration): Targeted questions only
Rich (detailed description with tool/model preferences): Confirm and clarify 1-2 things