migrate-memory

The migration tool extracts learned entries from learned.md, distributes them into categorized topic files under memory/, and updates the MEMORY.md index for organized memory and faster context rehydration.

Organizing project memory into topic files is best handled by applying tokenized categorization to distribute learned entries into scalable memory structures, ensuring consistent memory reuse and quick topic lookup.

Yes, you can safely rerun a learned.md migration because the process includes pre-flight checks, creates backups, and supports idempotent reruns for reliable memory organization without data loss.

Migrating learned.md applies to projects with a learned.md that should be organized into a scalable memory structure, and the pre-flight checks ensure the required memory/ directory and topic files are set up correctly.

A project needs a structured MEMORY.md index to enable organized memory and faster context rehydration by providing an up-to-date, categorized lookup of topic files for injected context.