feat(monitoring): wire HealthCollector state-change events to workflow trigger + notification pipeline

[mrveiss]

Context7 Score: 73/100

Test query: "Implement a real-time monitoring task that triggers a notification when a specific Linux service enters a failed state."

Audit Findings

What EXISTS

• autobot-slm-backend/slm/agent/health_collector.py — polls systemd via systemctl, maps states (active/failed/crash-loop/inactive)
• autobot-backend/services/notification_service.py — 4 channels (email, Slack, webhook, in-app)
• autobot-backend/services/trigger_service.py — REDIS_PUBSUB trigger type ready
• docs/guides/realtime-monitoring-notifications.md — 1300-line guide with working example

What is MISSING

1. HealthCollector doesn't publish to Redis — detects state changes but never emits to autobot:services:{name}:state_change pub/sub channel
2. No SERVICE_FAILED event type in NotificationEvent enum — only workflow lifecycle events exist
3. No service monitoring workflow template — users must manually wire everything
4. No end-to-end example connecting HealthCollector → REDIS_PUBSUB trigger → notification

Acceptance Criteria

• HealthCollector publishes {"service": name, "prev_state": ..., "new_state": ..., "error_context": ...} to autobot:services:{name}:state_change on every state transition
• NotificationEvent.SERVICE_FAILED added with default template: "Service {service} entered {state} state"
• Workflow template: autobot-backend/workflow_templates/service_health_monitor.yaml — trigger: REDIS_PUBSUB on autobot:services:*:state_change, step: send notification
• Example: docs/examples/service_failure_monitoring.py — complete runnable demo
• docs/user/guides/workflows.md updated with "Monitor a Linux Service" section
• Tests: state transitions, Redis pub/sub publish, notification dispatch

Files to Touch

• autobot-slm-backend/slm/agent/health_collector.py
• autobot-backend/services/notification_service.py
• autobot-backend/workflow_templates/service_health_monitor.yaml (new)
• docs/examples/service_failure_monitoring.py (new)
• docs/user/guides/workflows.md

[github-actions]

PR #3415 (merged to Dev_new_gui) references this issue with a close keyword.

feat(monitoring): wire HealthCollector state-change events to workflow trigger + notification pipeline

If this issue is fully resolved, close it manually. If work remains, no action is needed.