Compose Metrics into Effect Pipelines

Guideline

Metrics compose with Effect. Use Metric.increment, Metric.trackDuration, and Effect.tap to add observability to your pipelines without altering the core success value or control flow. Observability stays at the edges.

Rationale

By composing metrics into pipelines, you avoid scattering mutable counters or manual timing logic throughout your codebase. Effect's declarative style lets you wrap operations with Metric.trackDuration(timer) or add Effect.tap(() => Metric.increment(counter)) without changing what the pipeline returns.

Good Example

import { Effect, Metric } from "effect"

const successCounter = Metric.counter("operations_success_total")
const stepTimer = Metric.timer("step_duration")

const fetchUser = Effect.succeed({ id: 1, name: "Alice" }).pipe(
  Effect.delay("20 millis"),
  Metric.trackDuration(stepTimer)
)

const program = fetchUser.pipe(
  Effect.tap(() => Metric.increment(successCounter)),
  Effect.tap((user) => Effect.sync(() => console.log(user)))
)

Effect.runPromise(program)

Explanation: Metrics wrap and tap the pipeline. The success value flows through unchanged; counters and timers record observability data.

Anti-Pattern

Mixing business logic with metric recording (e.g., mutating counters inside the main Effect), or not using Effect.tap to keep metrics side-effect-only.