Codebase Agent Structure Generator

Overview

This agent is designed to systematically analyze an entire codebase and generate the foundational AGENTS.md structure required for advanced, context-aware AI coding agents. It enforces a 'nearest-wins' hierarchy, ensuring that agents receive highly relevant, token-efficient guidance rather than monolithic documentation.

Capabilities

• Repository Mapping: Determines if the project is a monorepo, multi-package setup, or simple single project.
• Technology Stacking: Identifies primary languages, frameworks, and key tools in use.
• Hierarchical Structuring: Creates a lightweight root AGENTS.md that points to detailed sub-guides within specific directories (e.g., apps/, services/).
• Pattern Extraction: Documents critical conventions, naming patterns, build systems, and testing setups for consistency.
• Token Efficiency Focus: Prioritizes actionable paths, globs, and commands over exhaustive content dumps.

Example Use Cases

1. Onboarding New Agents: Run this agent first to create the entire governance structure before any coding tasks begin.
2. Refactoring Large Codebases: Use it to map out which sub-systems need dedicated documentation guides when migrating or refactoring major sections.
3. Establishing Best Practices: Ensure all future AI interactions adhere to documented organizational patterns by generating the canonical guide files.