What problem does it solve? AI coding agents and new team members lack durable project context: coding standards, stack decisions, daily commands, and area-specific conventions live nowhere or get lost. This Skill bootstraps that context by scanning a codebase (or asking on greenfield projects) and writing tool-agnostic MEMORY.md files that every later skill and AI tool reads. ## Core Features & Use Cases - Phase-based routing: Detects project state (greenfield, undocumented brownfield, partially documented, or single area) and runs the matching audit mode, including legacy context file migration with permission. - Root and nested context files: Writes a short global root MEMORY.md plus nested area MEMORY.md files only where distinct conventions warrant them, never overwriting curated content. - Greenfield standards capture: Asks coding standards questions (architecture style presets like Clean Architecture, DDD, Functional, SOLID OOP) and seeds root MEMORY.md from the architecture spec and scope header. - Use Case: After scaffolding a new Laravel or Next.js app, run the audit to record the stack, commands, and chosen conventions in MEMORY.md so every subsequent AI-assisted task follows them; later, run it on src/auth to document that area's gotchas. ## Quick Start Ask the agent to run the audit skill on this project to create the MEMORY.md context files, optionally passing an area path like src/auth to document just that module.
