Operational Playbooks

# Operational Playbooks

## Daily Operations

### Morning Check-in (9:00 AM)
1. **Review Queue**
   - Check pending comments/posts
   - Approve or edit content
   - Regenerate low-quality suggestions

2. **Monitor Analytics**
   - Review previous day's performance
   - Check engagement metrics
   - Identify top-performing content

3. **Adjust Strategy**
   - Update target subreddits if needed
   - Modify posting frequency based on performance
   - Update templates based on feedback

### Midday Monitoring (1:00 PM)
1. **Check Agent Status**
   - Verify all agents are running
   - Check for any errors or warnings
   - Review ban detection alerts

2. **Content Review**
   - Approve high-priority content
   - Update library with successful templates
   - Reject inappropriate content

3. **Security Check**
   - Verify proxy rotation is working
   - Check fingerprint spoofing status
   - Review session health

### Evening Wrap-up (6:00 PM)
1. **Performance Analysis**
   - Review daily metrics
   - Export analytics reports
   - Update content library

2. **System Maintenance**
   - Check for system updates
   - Clean up old logs
   - Backup configuration

3. **Planning Tomorrow**
   - Adjust strategy based on today's results
   - Update content templates
   - Plan for upcoming events/topics

## Weekly Operations

### Monday: Strategy Review
1. **Weekly Performance Analysis**
   - Compare this week to last week
   - Identify trends and patterns
   - Adjust agent configurations

2. **Content Library Update**
   - Add successful templates
   - Remove underperforming content
   - Organize by topic/category

3. **Goal Setting**
   - Set weekly objectives
   - Adjust posting targets
   - Plan content themes

### Wednesday: Midweek Checkpoint
1. **Progress Assessment**
   - Evaluate goal progress
   - Identify obstacles
   - Adjust tactics as needed

2. **Agent Optimization**
   - Fine-tune behavior settings
   - Update subreddit lists
   - Test new content approaches

3. **Community Engagement**
   - Review feedback
   - Identify new opportunities
   - Plan engagement initiatives

### Friday: Weekly Wrap-up
1. **Performance Reporting**
   - Generate weekly reports
   - Export data for stakeholders
   - Document key learnings

2. **System Maintenance**
   - Update software components
   - Clean up temporary files
   - Review security settings

3. **Next Week Planning**
   - Set upcoming goals
   - Prepare content calendar
   - Schedule maintenance windows

## Monthly Operations

### Month-End Review
1. **Comprehensive Performance Analysis**
   - Compare to previous months
   - Identify seasonal trends
   - Evaluate ROI

2. **Agent Portfolio Management**
   - Review all active agents
   - Retire underperforming agents
   - Plan new agent deployments

3. **Technology Updates**
   - Update all software components
   - Test new LLM models
   - Implement infrastructure improvements

## Troubleshooting Playbooks

### Agent Not Posting
1. **Check Agent Status**
   ```bash
   # Verify agent is enabled
   docker-compose logs mcp-backend
   
   # Check configuration
   cat mcp_core/config.yaml
   ```

2. **Verify Credentials**
   - Check Reddit credentials in `.env`
   - Verify account is not suspended
   - Confirm app permissions

3. **Review Queue**
   - Check for pending items
   - Verify content approval status
   - Regenerate content if needed

### LLM Performance Issues
1. **Check Model Status**
   ```bash
   # Verify models are loaded
   curl http://localhost:11434/api/tags
   
   # Test model response
   ollama run qwen3-coder "Test prompt"
   ```

2. **Adjust Parameters**
   - Lower temperature for consistent output
   - Increase max tokens for longer content
   - Try alternative models

3. **Resource Management**
   - Check GPU/CPU usage
   - Restart Ollama service if needed
   - Consider upgrading hardware

### Security/Ban Detection
1. **Immediate Actions**
   - Pause all agents
   - Check account status on Reddit
   - Review recent activity

2. **Investigation**
   ```bash
   # Check security logs
   docker-compose logs mcp-backend | grep -i security
   
   # Review recent actions
   cat logs/security.log
   ```

3. **Recovery Plan**
   - Rotate proxies immediately
   - Update browser fingerprints
   - Implement longer cooldown periods
   - Review content guidelines

### Browser Automation Failures
1. **Check Browser Status**
   ```bash
   # Verify browser container
   docker-compose ps playwright-browser
   
   # Check browser logs
   docker-compose logs playwright-browser
   ```

2. **Restart Services**
   ```bash
   # Restart browser container
   docker-compose restart playwright-browser
   
   # Reinstall browsers if needed
   docker-compose exec playwright-browser npx playwright install chromium
   ```

3. **Configuration Review**
   - Check headless mode settings
   - Verify proxy configurations
   - Review browser profile settings

## Maintenance Playbooks

### Routine Maintenance (Weekly)
1. **System Updates**
   ```bash
   # Update Docker images
   docker-compose pull
   
   # Update Ollama models
   ollama pull qwen3-coder
   ```

2. **Log Management**
   ```bash
   # Rotate logs
   logrotate /etc/logrotate.d/mcp-agent
   
   # Clean old logs
   find logs/ -name "*.log" -mtime +30 -delete
   ```

3. **Backup Configuration**
   ```bash
   # Backup config files
   cp mcp_core/config.yaml backups/config-$(date +%Y%m%d).yaml
   
   # Export content library
   docker-compose exec mcp-backend python scripts/export_library.py
   ```

### Major Maintenance (Monthly)
1. **Infrastructure Review**
   - Check hardware health
   - Update system software
   - Review security patches

2. **Performance Optimization**
   - Analyze resource usage
   - Optimize database queries
   - Tune LLM parameters

3. **Feature Updates**
   - Deploy new features
   - Update documentation
   - Train team on changes

## Emergency Procedures

### System Outage
1. **Immediate Response**
   - Notify stakeholders
   - Document incident start time
   - Begin troubleshooting

2. **Recovery Steps**
   ```bash
   # Check all services
   docker-compose ps
   
   # Restart failed services
   docker-compose restart [service]
   
   # Check system resources
   docker stats
   ```

3. **Post-Incident Review**
   - Document root cause
   - Implement preventive measures
   - Update runbooks

### Security Breach
1. **Containment**
   - Isolate affected systems
   - Change all credentials
   - Disable compromised accounts

2. **Investigation**
   - Review security logs
   - Identify breach vector
   - Assess data impact

3. **Recovery**
   - Restore from clean backups
   - Implement additional security measures
   - Conduct security audit

### Data Loss
1. **Assessment**
   - Determine extent of data loss
   - Identify recovery options
   - Prioritize critical data

2. **Recovery**
   ```bash
   # Restore from backup
   cp backups/latest-backup.tar.gz /tmp/
   tar -xzf /tmp/latest-backup.tar.gz -C /
   
   # Verify data integrity
   python scripts/verify_data.py
   ```

3. **Prevention**
   - Implement automated backups
   - Test recovery procedures
   - Review backup strategy