How does it decide what information to prune?

It identifies completed todos, stale context that is no longer relevant to active projects, and duplicate information across different memory files.

Does this skill require manual input during operation?

No, it is designed to run in the background without user interaction, making decisions autonomously and noting any uncertainties in a journal for later review.

What is the primary purpose of the Memory Maintenance skill?

It acts as an automated cleanup crew for Claude's memory, ensuring that information stays current, state files are updated based on recent work, and outdated data is pruned.

How often should Memory Maintenance be run?

It is optimized for daily use, specifically as an 'end of day' task to process the day's conversations and prepare the environment for the next session.

Which files does this skill typically update?

It updates core state files like today.md, workspace.md, and human.md, as well as specific entity files within the people and projects directories.

Memory Maintenance

Name: Memory Maintenance
Author: ascorbic

byascorbic

•

108

•

Productivity & Workflow

Automates background maintenance of Claude's persistent memory by distilling conversations, updating state files, and pruning stale information.

The Memory Maintenance skill is a core component of the Macrodata framework, designed to keep Claude's persistent memory accurate, organized, and high-performing. It operates as an autonomous background process at the end of each day to distill facts from recent interactions, update core state files (such as workspace status and human preferences), and prune outdated entries. By managing search indexes and logging activities into a maintenance journal, this skill ensures that your AI's context remains relevant and optimized for future sessions without requiring manual user intervention.

Key Features

01Intelligent pruning of completed tasks and redundant context

02Self-managing search index maintenance for memory and conversations

03108 GitHub stars

04Automatic updates to workspace, project, and user preference files

05Autonomous end-of-day memory distillation and facts extraction

06Background execution designed to minimize user interaction and decision fatigue

Use Cases

01Ensuring AI memory indexes stay optimized for fast, relevant retrieval during active development

02Synchronizing daily conversation insights with persistent entity files for people and projects

03Maintaining long-term project context by archiving finished tasks and updating active threads overnight

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ascorbic/macrodata memory-maintenance

For use in Claude.ai and ChatGPT

Download Skill