Memory Keeper (Proactive Context Maintenance)

What problem does it solve?

This Skill replaces passive "context is stale" warnings with proactive, automated memory maintenance. Users no longer need to manually check or trigger context updates, eliminating mental overhead and ensuring the AI always has fresh information.

Core Features & Use Cases

• Proactive Detection: Automatically identifies when memory sync is needed (e.g., context >2 hours old, new commits, task completion, phase transitions, constitutional violations, quality gate failures).
• Intelligent Suggestions: Provides detailed reasons for sync and specific actions to be taken, with user confirmation.
• Automated Execution: Upon confirmation, it executes /memory-sync to update active_context.md, tasks_plan.md, lessons-learned.md, and other memory files.
• Use Case: Imagine you're deep in coding. Instead of a passive warning that your context is 5 hours old, Memory Keeper detects new commits and a task completion, then proactively asks, "Task T005 complete! Sync memory to capture learnings?" – you confirm, and it handles the rest.

Quick Start

Memory Keeper operates proactively in the background.

To trigger a manual check or confirm a suggestion:

"Sync memory" "Update context" "Capture learnings"