store: CT 300 claude-runner: Two-Tier Health Check Architecture

graph/configurations/ct-300-claude-runner-two-tier-health-check-architecture-061011.md

@@ -0,0 +1,43 @@
+---
+id: 06101183-a78b-4852-86eb-cae5557ace8c
+type: configuration
+title: "CT 300 claude-runner: Two-Tier Health Check Architecture"
+tags: [ct300, claude-runner, server-diagnostics, health-check, n8n, monitoring, claude-home, infrastructure, configuration]
+importance: 0.8
+confidence: 0.8
+created: "2026-03-01T14:16:30.152254+00:00"
+updated: "2026-03-01T14:16:30.152254+00:00"
+---
+
+# CT 300 claude-runner: Two-Tier Health Check Architecture
+
+## System Overview
+The headless Claude server diagnostics system runs on **CT 300** (hostname: `claude-runner`, IP: `10.10.0.148`). SSH alias is `claude-runner`.
+
+## Architecture
+Two-tier system:
+
+### Tier 1: health_check.py (Pure Python)
+- Triggered by **N8N on CT 210** on a schedule
+- No Claude CLI dependency — runs fine standalone
+- Checks **9 servers** via SSH:
+  - arr-stack, gitea, uptime-kuma, n8n, ubuntu-manticore, strat-database, pihole1, sba-bots, foundry
+- Checks performed: container status, HTTP endpoints, disk/memory/load
+- Exit codes:
+  - `0` = healthy
+  - `1` = auto-remediated
+  - `2` = escalate to Claude (Tier 2)
+
+### Tier 2: client.py (Claude Escalation)
+- Intended for deeper diagnosis when Tier 1 returns exit code 2
+- **Currently non-functional**: Claude CLI is NOT installed on CT 300
+- Would require npm/node + Claude CLI installation to enable
+
+## Config Location
+`~/.claude/skills/server-diagnostics/config.yaml`
+
+## Notifications
+Discord webhooks for alerts/notifications.
+
+## Gap
+Tier 2 cannot function until Claude CLI is installed on CT 300. See related memory on this gap.