What problem does it solve? Developers need specialized AI assistants for recurring tasks like code review or debugging, but configuring isolated agents with custom system prompts and correct file placement is error-prone without guidance. ## Core Features & Use Cases - Subagent File Authoring: Guides creation of .md files with required YAML frontmatter fields (name, description) and a markdown body that becomes the system prompt. - Scope Management: Explains project-level (.cursor/agents/) versus user-level (~/.cursor/agents/) placement and priority resolution when names collide. - Ready-Made Templates: Provides complete examples for a code reviewer, debugger, and data scientist subagent with effective delegation descriptions. - Use Case: After repeatedly asking the AI to review code, create a code-reviewer subagent with a proactive description so the AI automatically delegates review tasks to it. ## Quick Start Create a code-reviewer subagent in my project's .cursor/agents directory that proactively reviews code for quality and security.