memory-hygiene

Audit and prune junk memories in Clawdbot's LanceDB vector store.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/zkksdk/skills --skill memory-hygiene-zkksdk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-hygiene
Source: https://github.com/zkksdk/skills/tree/main/skills/memory-hygiene
Command: npx skills add https://github.com/zkksdk/skills --skill memory-hygiene-zkksdk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audit, clean, and optimize Clawdbot's vector memory (LanceDB). Use when memory is bloated with junk, token usage is high from irrelevant auto-recalls, or setting up memory maintenance automation.

Core Features & Use Cases

  • Audit memory to identify what's stored and identify junk memories.
  • Wipe and reseed memory to reset LanceDB with key facts, then restart the gateway for changes to take effect.
  • Configure memory capture and recall behavior to reduce future junk and automate maintenance tasks.
  • Use case: When memory bloat slows recall or token usage spikes, run memory hygiene to prune useless items and preserve critical contexts.

Quick Start

Audit memory to identify junk, wipe the LanceDB store, reseed with essential facts, and restart the gateway to apply changes.

Frequently Asked Questions about memory-hygiene

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

FAQPage Schema
How do I audit and clean up vector memory in LanceDB to reduce token waste?

You can audit and clean LanceDB vector memory by identifying junk memories, pruning useless items, wiping the store, and reseeding with essential facts to reduce token waste from irrelevant auto-recalls.

What causes high token usage from irrelevant auto-recalls in vector memory?

High token usage from auto-recalls occurs when vector memory becomes bloated with junk memories, causing the system to retrieve irrelevant contexts and consume excess tokens during processing.

How do I reseed LanceDB with essential facts after wiping the vector store?

Reseed LanceDB by wiping the existing store, adding clear and concise entries for critical memories, and restarting the gateway to ensure changes persist across restarts and take effect.

Can I configure autoCapture settings to prevent junk memories from bloating vector memory?

Yes, you can configure autoCapture settings to control memory capture and recall behavior, which reduces future junk and automates maintenance tasks to keep vector memory lean.

Does wiping the LanceDB vector store require a gateway restart to apply changes?

Yes, wiping and reseeding the LanceDB vector store requires a gateway restart to apply changes and ensure that the reset memory configuration persists across future restarts.