promote

To convert recurring patterns into permanent rules, the skill promotes successful behaviors from ephemeral auto-memory notes into persistent project-wide CLAUDE.md files or scoped .claude/rules/ directories. It distills descriptive memory entries into imperative instructions for consistent AI enforcement.

The best way to stop repeating corrections is to graduate the correction pattern into a permanent rule. The skill distills recurring auto-memory notes regarding API error responses into concise, imperative instructions inside CLAUDE.md, ensuring the behavior is enforced automatically.

Yes, you can scope documentation rules to a directory. Based on pattern scope and file path requirements, the skill targets either project-wide CLAUDE.md files for global instructions or scoped .claude/rules/ directories for localized, context-specific rule enforcement.

Auto-memory pattern graduation works by moving successful lessons learned from descriptive MEMORY.md notes into permanent, enforced project documentation. It refines these ephemeral notes into maintainable imperative rules, solving the problem of fragmented knowledge across your codebase.

Yes, you need existing auto-memory notes. The skill facilitates the distillation of descriptive memory notes into imperative instructions, meaning it requires recurring patterns previously captured in your project memory to function as input for rule generation.