Skip to content

Operations Overview

Day-to-day operational procedures for Project Aegis.

Operational Philosophy

OODA Loop

For every significant action: 1. Observe: Gather context (read files, check state) 2. Orient: Analyze against goals, identify constraints 3. Decide: Choose action, document reasoning 4. Act: Execute, log result, verify outcome

Three-Strike Debug Protocol

  1. Strike 1: Retry with modified approach
  2. Strike 2: Switch to local reasoning model
  3. Strike 3: STOP. Document. Escalate to human.

Work Priorities

  1. SHIP & ACTIVATE (70%): Wire features, fix bugs, validate
  2. REACTIVE WORK (20%): Discord tasks, Beads issues
  3. PROACTIVE WORK (10%): Research, documentation

Daily Schedule

Time (UTC) Activity
00:00-06:00 Maintenance window (backups, updates)
02:00 Memory consolidation job
03:00 Transcript digestion job
06:00-08:00 Morning check (status, planning)
08:00-22:00 Active work
22:00-24:00 Evening summary

Routine Procedures

Morning Routine (/morning)

  1. Check system health
  2. Review pending Beads tasks
  3. Query knowledge graph for context
  4. Verify wallet balance
  5. Post status to Discord
  6. Generate priority list

Evening Routine (/evening)

  1. Summarize completed tasks
  2. Record decisions made
  3. Note errors encountered
  4. Extract lessons learned
  5. Set tomorrow's priorities
  6. Post summary to Discord

Monitoring

Health Checks

# System status
curl https://aegisagent.ai/health

# Docker containers
docker ps

# Database
psql -h localhost -U agent -d aegis -c "SELECT 1;"

# Memory usage
free -h
df -h

Alert Channels

  • Discord #alerts: System health issues
  • Telegram: Time-sensitive alerts
  • WhatsApp: Emergency escalations

Scheduled Jobs

Schedule Job Purpose
Every 5 min Health check System monitoring
Every 30 min Discord poll Check #tasks
Hourly Balance check Verify wallet
02:00 UTC Memory consolidation Optimize storage
03:00 UTC Transcript digestion Update knowledge graph

Documentation

Last Updated: 2026-01-25