api-health-monitoring

Files (1)

• SKILL.md

SKILL.md (first lines)

# API Monitoring Skill

Design complete observability stacks for any API: health checks, metrics, alerting, and dashboards.

---

## Health Check Endpoints

### Liveness check — is the process alive?
```
GET /health/live
Response 200: { "status": "ok" }
Response 503: { "status": "error", "reason": "OOM" }
```

### Readiness check — can it serve traffic?
```
GET /health/ready
Response 200:
{
  "status": "ready",
  "checks": {
    "database": "ok",
    "cache": "ok",
    "message_queue": "ok",
    "external_api": "degraded"
  }
}
Response 503: { "status": "not_ready", "checks": { "database": "error" } }
```

### Deep health — full dependency tree
```
GET /health/deep
Response 200:
{
  "status": "healthy",
  "version": "2.1.0",
  "uptime_seconds": 86400,
  "dependencies": {
    "postgres": { "status": "ok", "latency_ms": 2 },
    "redis": { "status": "ok", "latency_ms": 0.5 },
    "stripe": { "status": "ok", "latency_ms": 120 }
  }
}
```

---

## SLI / SLO / SLA Definitions

| Metric | SLI (what to measure) | SLO (target) | SLA (committed) |
|--------|-----------------------|--------------|-----------------|
| Availability | % of successful requests | 99.95% | 99.9% |
| Latency | p99 response time | < 500ms | < 1000ms |
| Error rate | % 5xx responses | < 0.1% | < 0.5% |
| Throughput | requests per second | > 1000 rps | > 500 rps |

---

## Prometheus Metrics to Expose

```
GET /metrics  (prometheus scrape endpoint)

# Request counters
http_requests_total{method, route, status_code}
http_request_duration_seconds{method, route} (histogram)

# Business metrics
api_active_users_total
api_db_query_duration_seconds{query_type}
api_cache_hit_ratio
api_queue_depth{queue_name}

# Error metrics
api_errors_total{error_type, route}
api_circuit_breaker_state{service}
```

---

## Alerting Rules

```yaml
# Critical — page immediately
- alert: HighErrorRate
  expr: rate(http_requests_total{status_code=~"5.."}[5m]) / rate(http_requests_total[5m]) > 0.01
  for: 2m
  labels: { severity: critical }
  annotations: { summary: "Error rate > 1%" }

- alert: APIDown
  expr: up{job="api"} == 0
  for: 1m
  labels: { severity: critical }

- alert: HighLatency
  expr: histogram_quantile(0.99, rate(http_request_duration_seconds_bucket[5m])) > 1.0
  for: 5m
  labels: { severity: warning }

# Warning — Slack notification
- alert: DatabaseSlow
  expr: api_db_query_duration_seconds{quantile="0.95"} > 0.5
  for: 10m
  labels: { severity: warning }
```

---

## Structured Log Format (JSON)

```json
{
  "timestamp": "ISO8601",
  "level": "INFO|WARN|ERROR",
  "service": "api",
  "version": "2.1.0",

From LambdaTest/agent-skills, MIT.

More from LambdaTest/agent-skills