create-subagent

Create Cursor subagents with Markdown files and YAML frontmatter.

2|Updated Apr 25, 2026
One-click install
npx skills add https://github.com/caezium/skills --skill create-subagent-caezium
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-subagent
Source: https://github.com/caezium/skills/tree/main/create-subagent
Command: npx skills add https://github.com/caezium/skills --skill create-subagent-caezium

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you create and manage specialized subagents that run in isolated contexts with tailored system prompts, making complex AI workflows modular and reusable.

Core Features & Use Cases

  • Frontmatter-driven configuration: Define each subagent with a YAML frontmatter file including name and description.
  • Location-based scoping: Use project-level (.cursor/agents/) or user-level (~/.cursor/agents/) subagents to share or personalize agents.
  • System-prompt-driven behavior: Write the system prompt in the body to control delegation, workflow, and outputs.
  • Use Case: Create a code-reviewer subagent to enforce coding standards and a debugger subagent to guide root-cause analyses.

Quick Start

Create a new subagent by adding a Markdown file with YAML frontmatter to the appropriate agents directory and filling the body with the system prompt.

Frequently Asked Questions about create-subagent

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

FAQPage Schema
How do I create custom subagents for Cursor?

To create custom subagents for Cursor, you add Markdown files with YAML frontmatter defining the name and description, then write the system prompt in the body to control delegation and workflow.

What is the difference between project-level and user-level subagents?

Project-level subagents are stored in `.cursor/agents/` for project-specific sharing, while user-level subagents use `~/.cursor/agents/` to personalize agents across all your projects.

When should I use isolated subagents in my AI workflow?

Use isolated subagents to make complex AI workflows modular and reusable, such as creating a code-reviewer to enforce standards or a debugger to guide root-cause analysis independently.

How do I configure subagent behavior using YAML frontmatter?

You configure subagent behavior by defining the name and description in YAML frontmatter at the top of the Markdown file, and writing the tailored system prompt in the body.

Does this subagent creation approach work without external dependencies?

Yes, this approach works without external dependencies because it relies entirely on placing Markdown files with YAML frontmatter into the correct local Cursor agents directories.

What are the limitations of managing subagents with Markdown files?

Limitations include manually ensuring correct directory placement and YAML syntax, as subagents will not function if the file structure is invalid or the system prompt lacks clear delegation instructions.