self-protection

name	self-protection
description	Self-protection system for OpenClaw agent to prevent frequent crashes, implement load balancing, hourly heartbeat self-check, and automatic recovery.
read_when	["Agent is experiencing frequent crashes","System needs load balancing","Hourly health check is required","Memory or process monitoring needed"]
metadata	{"clawdbot":{"emoji":"🛡️","requires":{"bins":["ps","top","uptime","curl"]}}}

Self Protection System

This skill provides self-protection mechanisms for OpenClaw agent to ensure stable operation and prevent frequent crashes.

Features

1. Load Balancing

• Monitor system resource usage
• Limit concurrent operations
• Queue management for heavy tasks
• Automatic throttling based on load

2. Hourly Heartbeat Self-Check

• System health monitoring
• Memory usage checks
• Process health verification
• External service connectivity tests

3. Crash Prevention

• Early warning detection
• Automatic recovery procedures
• Crash logging and analysis
• Fallback mechanisms

4. Performance Monitoring

• CPU usage tracking
• Memory usage monitoring
• Disk space checks
• Network connectivity verification

5. Automatic Recovery

• Failed operation retry logic
• Graceful degradation
• Resource cleanup
• State restoration

Quick Commands

Health Check

self-protection check

Load Monitoring

self-protection load

Resource Report

self-protection resources

Start Hourly Heartbeat

self-protection heartbeat start

Stop Hourly Heartbeat

self-protection heartbeat stop

Crash Analysis

self-protection crash-log

Recovery Procedures

self-protection recover

Implementation Details

Load Balancing Algorithm

1. Monitor system load using uptime, top, ps
2. Track concurrent operations count
3. Implement operation queue with priority
4. Throttle new requests when load > 70%

Heartbeat Check Schedule

1. Run hourly (via cron or scheduler)
2. Check:
   • System uptime and stability
   • OpenClaw service status
   • Memory usage and limits
   • External dependencies (API connectivity)
3. Log results to health log file
4. Alert if thresholds exceeded

Crash Prevention Measures

1. Rate limiting for external API calls
2. Memory usage monitoring (set thresholds)
3. Process watchdog for critical services
4. Graceful shutdown on signal
5. Automatic cleanup of orphaned processes

Performance Thresholds

• CPU usage: >80% triggers throttling
• Memory usage: >85% triggers cleanup
• Disk space: <10% free triggers alert
• Network: latency >500ms triggers fallback

Implementation Files

The skill includes:

1. scripts/self-protection.sh - Main protection script
2. scripts/load-balancer.sh - Load balancing algorithm
3. scripts/heartbeat.sh - Hourly heartbeat check
4. scripts/recovery.sh - Crash recovery procedures
5. references/thresholds.json - Resource thresholds
6. references/crash-log.md - Crash analysis log
7. cache/health-status.json - Current health status
8. cache/system-metrics.log - System metrics history

Usage Examples

Basic Health Check

self-protection check

Start Hourly Monitoring

self-protection heartbeat start

Load Balancing Setup

self-protection load setup

Crash Recovery

self-protection recover --force

Analyze Recent Crash

self-protection crash-log --last

Integration with OpenClaw

This skill integrates with:

• OpenClaw's own health monitoring
• System cron for heartbeat scheduling
• Memory monitoring tools
• Process management utilities

Installation Requirements

• ps - Process monitoring
• top - System load monitoring
• uptime - System uptime
• curl - API connectivity tests
• jq - JSON processing (optional)
• awk - Text processing (optional)