memory-maintenance

Compact, archive, and validate the .agent-memory/ store.

Updated Mar 20, 2026
One-click install
npx skills add https://github.com/dynamic-dome/agentic-os --skill memory-maintenance-dynamic-dome
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-maintenance
Source: https://github.com/dynamic-dome/agentic-os/tree/main/skills/memory-maintenance
Command: npx skills add https://github.com/dynamic-dome/agentic-os --skill memory-maintenance-dynamic-dome

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill automates the maintenance and organization of the .agent-memory/ store, ensuring the integrity and accessibility of project knowledge while preventing bloat and inefficiency.

Core Features & Use Cases

  • Memory Compaction: Regularly compact and verify the .agent-memory/ store to maintain its integrity.
  • Data Archiving: Archive old data based on defined thresholds to prevent bloat and improve performance.
  • Stale Pattern Pruning: Remove outdated patterns that are no longer relevant.
  • Session Summary Length Control: Enforce length limits on session summaries to keep them concise.
  • Learnings Deduplication: Deduplicate learnings entries to maintain clarity and prevent redundancy.
  • Consistency Checks: Perform cross-file integrity checks to ensure data consistency.
  • Use Case: Use this Skill to ensure that your project's knowledge base remains up-to-date and accessible, saving time and reducing the risk of outdated information.

Quick Start

To initiate memory maintenance, simply say "clean memory" or "memory maintenance".

Frequently Asked Questions about memory-maintenance

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate memory maintenance and prevent project knowledge bloat?

Automate memory maintenance using threshold-based archiving and stale pattern pruning to prevent project knowledge bloat. This process validates JSON data, deduplicates learnings, and cleans up working files to ensure knowledge remains accessible and efficient.

When should I run data archiving and compaction on my project knowledge store?

Run data archiving and compaction during or after a session wrap-up. Executing memory maintenance at this stage ensures threshold-based archiving and scratch file cleanup accurately capture the latest session data without interrupting active workflows.

What is the best way to prune stale patterns and deduplicate learnings in agent memory?

The best way to prune stale patterns and deduplicate learnings is through automated cross-file integrity checks. These consistency checks identify outdated patterns and redundant entries, maintaining clarity while enforcing session summary length limits.

Does memory maintenance handle JSON validation and cross-file consistency checks?

Yes, memory maintenance handles JSON validation and cross-file consistency checks. It performs automated integrity checks across the .agent-memory/ store to ensure data consistency and verify that archived project knowledge remains structurally valid.

How do I clean up working and scratch files during session wrap-up?

Clean up working and scratch files during session wrap-up by triggering automated memory maintenance. This process enforces session summary length limits and removes temporary scratch files, ensuring the project knowledge store remains uncluttered.