What problem does it solve?
This Skill addresses the challenge of maintaining a clean, relevant, and efficient AI memory by periodically auditing stored information, identifying stale or redundant records, and consolidating weak signals to improve overall AI learning and decision-making.
Core Features & Use Cases
- Memory Audit: Scans memory directories for stale records, weak patterns, near-duplicates, and high-error domains.
- Record Management: Identifies records for pruning, consolidation, or confidence boosting based on defined criteria.
- User Approval Workflow: Ensures all modifications to memory are explicitly approved by the user.
- Calibration Data: Generates domain error density reports to inform future skill loading.
- Use Case: After a complex multi-task specification, run
/reflect to ensure the AI's memory is up-to-date and free of outdated information, preventing it from acting on stale knowledge in future sessions.
Quick Start
Run the reflect skill to audit the AI's memory for stale records and suggest improvements.