v1.2.0 - Agent Resilience and Crash Recovery

Overview

This release adds comprehensive agent process resilience features to prevent and recover from AI CLI crashes, particularly Claude Code memory leaks that can reach 13-120GB+ in extended sessions.

Key Features

🛡️ Agent Resilience Phase

New bootstrap phase 8 installs crash prevention and recovery tools:

Feature	Description
Memory Watchdog	systemd service that warns at 8GB, kills Claude at 13GB
cgroups Limiting	run-claude-limited enforces hard memory limits
Session Persistence	agent-session tmux wrapper with auto-reattach
Trend Analysis	vm-health-check --trend predicts OOM 30-60 min ahead

📊 Resource Planning

• Default RAM: 8GB → 16GB
• Default swap: 4GB → 8GB
• New --memory and --vcpus flags for setup_cloud.sh
• New RESOURCES.md with parallel deployment guide

🧹 Cleanup

• Removed .mcp.json from git (local config)
• Removed vestigial guest/vm-health-check.sh
• Fixed README accuracy

New Commands (Inside VM)

vm-health-check              # Quick health with Claude memory tracking
vm-health-check --watch      # Continuous monitoring (30s refresh)
vm-health-check --trend      # Memory trend analysis with OOM prediction
run-claude-limited           # Run Claude with 12GB hard limit
run-claude-limited 8G        # Custom limit
agent-session                # Persistent tmux session
systemctl status claude-watchdog  # Check watchdog service

Recommended 64GB Host Configuration

# Golden image (16GB, stopped during parallel work)
./setup_cloud.sh --memory 16384 --vcpus 4

# 3 parallel clones at 12GB each
./clone_manager.sh create ubuntu2404-agent --linked --memory 12288 --vcpus 4
./clone_manager.sh create ubuntu2404-agent --linked --memory 12288 --vcpus 4
./clone_manager.sh create ubuntu2404-agent --linked --memory 12288 --vcpus 4

Full Changelog

See CHANGELOG.md for complete details.

---

Full Changelog: v1.1.0...v1.2.0