# Infrastructure Monitor

> Continuous infrastructure health monitoring for trading services. Discovers new metrics and alerts proactively.

- Skill: `knuckles-team/infrastructure-monitor` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/infrastructure-monitor`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/infrastructure-monitor/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/infrastructure-monitor

---

# Infrastructure Monitor Workflow (Cron: every 15 minutes)

## Workflow Execution Steps

### Step 1: container-health
Check container health for all trading services.
Tool: `portainer_docker(action="docker_list_containers")`

### Step 2: error-traces
Query recent error traces from observability.
Tool: `langfuse_observability(action="trace_list", tags=["finance"])`

### Step 3: system-resources
Check disk/memory/CPU on trading nodes.
Tool: `systems_manager` resource checks.

### Step 4: dns-check
Verify DNS resolution for all trading services.
Tool: `technitium_zones(action="get_records")`

### Step 5: kg-persist
Store HealthSnapshot node in KG.
Tool: `graph_write(action="add_node", node_type="HealthSnapshot", ...)`

### Step 6: discover
Proactively identify new metrics or alerts from infrastructure.
Analyze container logs and system metrics for anomalies.
Recommend new monitoring rules to add.

## Execution

Run this workflow as a dependency-ordered DAG. Steps with no unmet `depends_on` run in parallel; dependents run after their prerequisites complete.

- **Run first (in parallel):** Step 1 — container-health; Step 2 — error-traces; Step 3 — system-resources; Step 4 — dns-check; Step 5 — kg-persist; Step 6 — discover

**Execution:** If graph-os is reachable, offload the whole DAG via `graph_orchestrate action=execute_workflow` (or the `kg-delegate` skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet `depends_on` in parallel, then their dependents.