feat: add automated directory health monitor

[hermes-alby]

Summary

Add an automated weekly health monitor that runs every Monday at 9am UTC via cron job.

What it checks

URL Health — HTTP status checks for ALL apps (reads from apps.json at runtime)

• Detects dead domains, 4xx/5xx errors, timeouts
• New apps added via PR are automatically included

GitHub Activity — dynamically discovers and monitors open-source repos

• Priority discovery: explicit github field → GitHub URL → known mapping → description scan
• Flags repos as stale (>180 days) or dead (>365 days or archived)
• No hardcoded list to maintain — new open-source apps are auto-detected

Report format

The monitor outputs a structured report with:

• ✅ Healthy apps and repos
• ❌ Dead apps (with URL, status code, error)
• 🟡 Stale repos (>180 days since last push)
• 🔴 Dead repos (>365 days or archived)
• Detailed JSON report saved as monitor_report.json

Integration

• Installed as a cron job running weekly via Hermes agent
• If dead apps or archived repos are found, the agent automatically prepares a PR to remove them
• When adding a new app, just include a "github": "owner/repo" field to have it tracked