hmem-self-curate

Audit memory prefixes and prune stale or obsolete entries.

23|5|Updated Feb 19, 2026
One-click install
npx skills add https://github.com/Bumblebiber/hmem --skill hmem-self-curate
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hmem-self-curate
Source: https://github.com/Bumblebiber/hmem/tree/main/skills/hmem-self-curate
Command: npx skills add https://github.com/Bumblebiber/hmem --skill hmem-self-curate

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Keeps long-term memory clean, relevant, and navigable by auditing and curating entries, reducing noise and improving retrieval efficiency across sessions.

Core Features & Use Cases

  • Health checks identify broken links, orphaned entries, stale favorites, and broken obsolete chains.
  • Guided workflow to audit entries prefix by prefix and apply curation decisions, including marking obsolete or merging duplicates.
  • Bulk operations to update, relocate, or consolidate memory nodes while preserving historical references and traceability.

Quick Start

Run memory_health to identify issues, then review a chosen memory prefix to begin systematic curation.

Frequently Asked Questions about hmem-self-curate

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

FAQPage Schema
How do I clean up stale long-term memory entries and improve retrieval efficiency?

To clean stale long-term memory, you can run periodic self-curation workflows that audit memory prefixes, identify orphaned entries, and prune obsolete chains to reduce noise and improve retrieval reliability across sessions.

What is the best way to identify broken links and orphaned entries in a memory database?

Identifying broken links and orphaned entries requires running a memory health check, which audits stored memory nodes to detect stale favorites, broken obsolete chains, and invalid references for systematic curation.

How do I perform bulk operations to merge or relocate duplicate memory nodes?

Bulk operations to merge or relocate duplicate memory nodes use update_memory, append_memory, and move_memory commands to consolidate entries while preserving historical references and maintaining full traceability.

Can I audit specific memory prefixes instead of reviewing the entire memory database at once?

Yes, you can audit specific memory prefixes individually, enabling a guided workflow that reviews entries prefix by prefix to apply focused curation decisions like obsoleting or merging duplicates without processing the entire database.

What happens to historical data when I obsolete or merge memory entries during curation?

When obsoleting or merging memory entries during curation, historical references and traceability are preserved, ensuring that relocated or consolidated memory nodes maintain their audit trail and do not lose past context.

When should I run memory curation to prevent my long-term memory from becoming unreliable?

Memory curation should be run periodically as a maintenance task to prevent long-term memory from accumulating noise, ensuring stored entries remain clean, relevant, and navigable for efficient future retrieval.