rekal-hygiene

Scan rekal SQLite memory stores for duplicates, conflicts, and stale conversations.

52|4|Updated Apr 11, 2026
One-click install
npx skills add https://github.com/janbjorge/rekal --skill rekal-hygiene
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rekal-hygiene
Source: https://github.com/janbjorge/rekal/tree/main/skills/rekal-hygiene
Command: npx skills add https://github.com/janbjorge/rekal --skill rekal-hygiene

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Performs periodic maintenance and cleanup of a rekal memory database to surface duplicates, contradictions, stale conversations, and low-quality memories so the user can approve corrective actions without risking automatic deletions or modifications.

Core Features & Use Cases

  • Conflict detection and resolution proposals: Finds contradictory memories, classifies conflict types, and proposes supersede, keep-both, or link/remove actions for user approval.
  • Duplicate clustering and consolidation: Groups topic clusters, identifies the most complete version, and proposes superseding redundant entries while preserving history.
  • Stale conversation and quality audits: Flags empty or old conversations, short or mis-typed memories, and suggests updates, supersedes, or deletions subject to explicit consent.
  • Safe, user-approved execution: Never auto-deletes or auto-modifies; presents a consolidated action plan and executes only after explicit user approval. Scales by prioritizing high-impact items for large databases.

Quick Start

Run rekal-hygiene to scan the memory database for duplicates, conflicts, stale conversations, and quality issues, then present a proposed action plan for your approval.

Frequently Asked Questions about rekal-hygiene

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

FAQPage Schema
How do I clean up duplicate and contradictory memories in a SQLite database?

To clean up duplicate and contradictory memories in a SQLite database, run a hygiene scan to cluster topics, identify the most complete versions, and propose supersede or delete operations for explicit user approval.

How does memory conflict detection work for conversational data?

Memory conflict detection works by scanning stored conversational memories to classify contradiction types and proposing supersede, keep-both, or link/remove actions, which are withheld until explicit user approval.

Can I automatically delete stale conversations from my memory database?

You cannot automatically delete stale conversations from your memory database; the system flags empty, old, or low-quality memories and suggests updates or deletions, but executes actions only after explicit user consent.

What is the best way to perform periodic maintenance on a rekal memory database?

The best way to perform periodic maintenance on a rekal memory database is to run a hygiene scan that surfaces duplicates, conflicts, stale conversations, and quality issues, then review and approve the proposed consolidated action plan.

Does memory cleanup work for large databases with many projects and timelines?

Memory cleanup works for large databases by prioritizing high-impact items across projects and timelines, scaling the audit to surface significant duplicates, contradictions, and stale entries for user-approved corrective actions.

What are the limitations of automated memory hygiene and cleanup?

A limitation of automated memory hygiene is that it requires read access to the local SQLite database and never auto-modifies or auto-deletes entries, meaning all supersede, update, link, or delete operations require explicit user confirmation.