Metrics

from triform import metrics

def process_order(order):
    start = time.time()
    
    # Your logic
    result = fulfill_order(order)
    
    # Track custom metrics
    metrics.increment("orders_processed")
    metrics.gauge("order_value", order.total)
    metrics.histogram("processing_time", time.time() - start)
    metrics.counter("items_shipped", order.item_count)
    
    return result

metrics.increment("api_calls")
metrics.increment("successful_logins")
metrics.add("messages_sent", count=5)

metrics.gauge("queue_size", current_size)
metrics.gauge("active_users", user_count)
metrics.set("temperature", 72.5)

metrics.histogram("request_duration", duration_ms)
metrics.histogram("payload_size", size_bytes)
metrics.observe("response_time", elapsed)

metrics.summary("processing_time", duration)

metrics.increment("api_calls", tags={
    "endpoint": "/users",
    "method": "GET",
    "status": "200"
})

api_calls{endpoint="/users", status="200"}

alert:
  name: "High error rate"
  metric: error_rate
  condition: "> 5%"
  duration: "5 minutes"
  notify:
    - email: "oncall@example.com"
    - slack: "#alerts"

metrics.increment("orders_completed")
metrics.gauge("revenue_today", daily_revenue)
metrics.histogram("customer_satisfaction", rating)

metrics.histogram("api_latency", response_time)
metrics.gauge("memory_usage", memory_mb)
metrics.increment("cache_hits")
metrics.increment("cache_misses")

metrics.gauge("daily_llm_cost", llm_cost)
metrics.counter("api_calls_by_endpoint", tags={"endpoint": endpoint})

metrics.gauge("agent_response_quality", quality_score)
metrics.histogram("user_satisfaction", rating)
metrics.increment("thumbs_up")
metrics.increment("thumbs_down")

metrics.gauge("active_projects", count)
metrics.gauge("executions_per_hour", rate)

from triform import Metrics

# Record metrics
Metrics.increment("orders", tags={"region": "us-west"})
Metrics.gauge("queue_depth", 42)
Metrics.histogram("latency", 123.45)

# Query metrics
data = Metrics.query(
    metric="api_latency",
    aggregation="p95",
    start_time="2025-10-01",
    end_time="2025-10-02",
    filters={"endpoint": "/users"}
)

# Create dashboard
dashboard = Metrics.create_dashboard(
    name="Production Overview",
    widgets=[
        {"type": "line", "metric": "request_rate"},
        {"type": "gauge", "metric": "error_rate"},
        {"type": "bar", "metric": "endpoint_usage"}
    ]
)

metrics_export:
  provider: prometheus
  endpoint: https://prometheus.example.com
  scrape_interval: 15s

metrics.increment("high_volume_event", sample_rate=0.1)
# Records 10% of events, multiplies by 10 for accurate count