marrs
Memory maintenance helper for any RAG/vector database. Includes save_memory() helper, monitor and defrag script templates, hot queue support, and configurable defaults. Declares 'requests' dependency. Fully generic, safe, and configurable once you edit the config. Accurate description of all contents.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/agenthyjack/marrsmarrs
Memory maintenance helper for any RAG/vector database.
Complete Setup Guide (so it works for any user)
- Install via ClawHub or copy the folder.
pip install requests(the only external dependency).- Edit
scripts/config.pywith your own RAG details (examples are placeholders only). - Create two scheduled jobs to run the monitor and defrag scripts (see your platform's cron/docs).
- Test with the example in the Basic Usage section below.
Review the three small Python scripts before use — they are short and easy to audit.
Configuration (scripts/config.py)
RAG_URL = "http://your-rag-server:port" # ← Replace with your own
DEFAULT_COLLECTION = "memory" # Change to your main collection
MONITOR_INTERVAL_SECONDS = 300
DEFRAG_INTERVAL_SECONDS = 86400
Basic Usage
from scripts.save_memory import save_memory
save_memory("Your content here", collection="your-collection")
What it contains
save_memory()helper that POSTs to your RAG/ingestendpoint- Template scripts for monitor and defrag (background maintenance)
- Hot queue logic for fast retrieval of frequent items
- Configurable defaults (you must edit them)
No hardcoded paths, no internal systems, no credentials.
Security Notes
- Only interacts with the RAG_URL you configure.
- Uses the 'requests' library (installed separately).
- Prints status messages to stdout (visible in your logs).
- The crons you create will run the scripts autonomously — only add them if you trust the code after review.
- Recommended: run the scripts in an isolated environment first.
This package is instruction + runnable scripts. The SKILL.md accurately describes everything it contains. No private data, no keys, no tokens, no names, no locations.
Version: 1.5.0 Status: Honest metadata, declared dependency, clear audit instructions. Clean for public use.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-agenthyjack-marrs": {
"enabled": true,
"auto_update": true
}
}
}Related Skills
dial-a-cron
Stateful cron system for OpenClaw with persistent memory, change detection, smart routing, token budget tracking, and self-healing. Requires 'openclaw' and 'gog' CLIs. Broad I/O capabilities (file reads, HTTP requests, shell execution). Must be reviewed carefully before use. See security notes below.
agent-health-diagnostics
Diagnose and fix the 4 most common OpenClaw agent failures — heartbeat spam, API rate limit cascades, channel death loops, and memory/embedding errors. Battle-tested across a 6-agent multi-host deployment.
multi-agent-status
Cross-agent health monitoring for multi-host OpenClaw deployments. Each agent pushes structured status reports (JSON) to a central location. A PM/monitoring agent reads them and alerts on failures. Works across Windows, Linux, and mixed environments.