Memory Hygiene

If your Clawdbot's vector memory is filling up with garbage like heartbeat messages and temporary status updates, this is the cleanup toolkit. It shows you how to audit what's actually stored in LanceDB, nuke everything when it gets out of hand, and reseed with only the facts that matter. The real value is the autoCapture config tip, turning off the firehose of automated memory writes that cause most of the bloat. Includes a cron job template for monthly wipes and practical guidelines on what deserves storage versus what's just noise. Straightforward maintenance hygiene for anyone running memory enabled agents long term.

skills.sh installs: 356
GitHub stars: 66
last updated: Jun 3, 2026

View on GitHub

Install

$ npx -y skills add aaaaqwq/claude-code-skills --skill memory-hygiene --agent claude-code

Installs into .claude/skills of the current project.

Files

Memory Hygiene

Keep vector memory lean. Prevent token waste from junk memories.

Quick Commands

Audit: Check what's in memory

memory_recall query="*" limit=50

Wipe: Clear all vector memory

rm -rf ~/.clawdbot/memory/lancedb/

Then restart gateway: clawdbot gateway restart

Reseed: After wipe, store key facts from MEMORY.md

memory_store text="<fact>" category="preference|fact|decision" importance=0.9

Config: Disable Auto-Capture

The main source of junk is autoCapture: true. Disable it:

{
  "plugins": {
    "entries": {
      "memory-lancedb": {
        "config": {
          "autoCapture": false,
          "autoRecall": true
        }
      }
    }
  }
}

Use gateway action=config.patch to apply.