When should I perform a memory review?

You should run a review when the dashboard indicates 'Yellow' (60-90% capacity) or 'Red' (>90% capacity) to prevent Claude from dropping older entries.

Is there a quick way to check status without a full report?

Yes, you can use the /si:status --brief command to get a one-line summary of your memory usage and health status.

How does it identify stale entries?

The skill scans your MEMORY.md file for file paths and verifies if those files still exist on your disk, flagging broken references for deletion.

What does the /si:status command do?

It generates a dashboard showing the line counts, topic files, capacity, and stale entries within your Claude memory system to ensure the AI isn't losing context.

Memory Health Dashboard

Name: Memory Health Dashboard
Author: alirezarezvani

byalirezarezvani

•

9,958

•

Analytics & Monitoring

Monitors and reports on the health, capacity, and integrity of Claude's auto-memory and project rule systems.

The Memory Health Dashboard provides a comprehensive diagnostic overview of your project's memory state by analyzing file sizes, rule counts, and link integrity across all memory systems. It helps developers manage Claude's context window efficiency by flagging stale entries, identifying capacity bottlenecks in MEMORY.md and CLAUDE.md, and offering actionable recommendations to maintain peak AI performance during complex, long-running coding sessions.

Key Features

01Real-time monitoring of MEMORY.md and CLAUDE.md line counts

029,958 GitHub stars

03Detailed auditing of local and global .claude rule files

04Capacity heatmaps (Green/Yellow/Red) to prevent memory loss

05Brief mode for quick, one-line status updates

06Automated stale reference detection for deleted files

Use Cases

01Preventing memory overflows and data loss during long development cycles

02Identifying and cleaning up broken file references in the memory system

03Optimizing project rules to ensure Claude maintains the most relevant context

Key Features

01Real-time monitoring of MEMORY.md and CLAUDE.md line counts

029,958 GitHub stars

03Detailed auditing of local and global .claude rule files

04Capacity heatmaps (Green/Yellow/Red) to prevent memory loss

05Brief mode for quick, one-line status updates

06Automated stale reference detection for deleted files

Use Cases

01Preventing memory overflows and data loss during long development cycles

02Identifying and cleaning up broken file references in the memory system

03Optimizing project rules to ensure Claude maintains the most relevant context