Key Focus: Distinguishing between manual configuration (claude.md) and automatic learning (memory.md).
Key Takeaways
The “Brain” Duality:
Explicit (claude.md): You tell Claude how to behave, what architecture to follow, and what the roadmap looks like.
Implicit (memory.md): Claude tells itself what it has learned about the project (Auto-Memory).
Roadmap-Driven Development: Using claude.md as a live tracker for feature status, allowing Claude to resume work with full knowledge of what is “DONE” vs “TODO”.
Auto-Memory: Claude identifies recurring patterns (e.g., currency handling, naming schemes) and saves them to memory.md to ensure cross-session consistency without user intervention.
Team Consistency: By committing the .claude folder, the “Implicit” and “Explicit” memory is shared across the whole team.