agentleFS
niieaniniieani/caara/.agents/skills/effect-ts/SKILL.md

effect-ts

Skill5 starsChanged 4 months agoMITView on GitHub

Write idiomatic Effect v4 TypeScript following official best practices from effect-solutions and the Effect source. Use when writing, reviewing, or refactoring Effect code: services (ServiceMap.Service), layers and dependency injection, error handling (Schema.TaggedErrorClass), data modeling (Schema.Class, branded types, variants), testing (@effect/vitest), HTTP clients (effect/unstable/http), CLI tools (effect/unstable/cli), config, observability, and project setup. Triggers on: 'Effect', 'effect-ts', '@effect/', 'Schema', 'ServiceMap', 'Layer', 'Effect.gen', 'Effect.fn', 'TaggedError', 'branded types', or any Effect-TS related code.

  • Reads credentials

17 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use shadcn, web-design-guidelines and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • DatabricksSkill

    shadcn

    Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", or "switch to --preset".

    Also in 3 projects that use this

  • VercelSkill

    web-design-guidelines

    Review UI code for Web Interface Guidelines compliance. Use when asked to "review my UI", "check accessibility", "audit design", "review UX", or "check my site against best practices".

    Also in 3 projects that use this

  • GoogleSkill

    grill-me

    A relentless interview to sharpen a plan or design.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 0 are distinctive, and 7 are very specific to this project.

  • 0

    Common

    Most of the 17 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 7

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Search source for implementations: grep -r "pattern" ./references/effect/packages/effect/src/

    line 14
  2. Very specific:

    Use Schema.Class for domain records.

    line 107
  3. Very specific:

    Use Schema.TaggedClass + Schema.Union for variants (OR types):

    line 131
  4. Very specific:

    Use Layer.sync for synchronous implementations, Layer.effect for effectful ones

    line 217
  5. Very specific:

    Use it.effect for Effect-based tests (provides TestContext with TestClock)

    line 234
  6. Very specific:

    Use it.live for real time / real clock

    line 235
  7. Very specific:

    Use it.layer only when sharing expensive resources across a suite

    line 237

“Common” is measured: kept by at least 6 in 10 of the 17 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Effect-TS (v4)

Patterns from effect-solutions and the Effect source. This covers the latest v4 APIs.

Local Source References

  • effect-solutions (best practices, docs, examples): ./references/effect-solutions/
  • effect monorepo (canonical source for all @effect/* packages): ./references/effect/
  • Search source for implementations: grep -r "pattern" ./references/effect/packages/effect/src/

Effect.gen and Effect.fn

Effect.gen provides sequential, readable composition (like async/await for Effect):

import { Effect } from "effect"

const program = Effect.gen(function* () {
  const data = yield* fetchData
  yield* Effect.logInfo(`Processing: ${data}`)
  return yield* processData(data)
})

Effect.fn adds call-site tracing and named spans. Use for all service methods:

const processUser = Effect.fn("processUser")(function* (userId: string) {
  yield* Effect.logInfo(`Processing user ${userId}`)
  const user = yield* getUser(userId)
  return yield* processData(user)
})

// Second argument for cross-cutting concerns (retry, timeout)
const fetchWithRetry = Effect.fn("fetchWithRetry")(
  function* (url: string) {
    const data = yield* fetchData(url)
    return yield* processData(data)
  },
  flow(
    Effect.retry(Schedule.recurs(3)),
    Effect.timeout("5 seconds")
  )
)

ServiceMap.Service

Define services as classes with a unique tag and typed interface:

import { Effect, ServiceMap } from "effect"

class Database extends ServiceMap.Service<
  Database,
  {
    readonly query: (sql: string) => Effect.Effect<unknown[]>
    readonly execute: (sql: string) => Effect.Effect<void>
  }
>()("@app/Database") {}

Implement with Layer.effect or Layer.sync, using Effect.fn for all methods:

import { Effect, Layer } from "effect"

class Users extends ServiceMap.Service<
  Users,
  {
    readonly findById: (id: UserId) => Effect.Effect<User, UserNotFoundError>
    readonly all: () => Effect.Effect<readonly User[]>
  }
>()("@app/Users") {
  static readonly layer = Layer.effect(
    Users,
    Effect.gen(function* () {
      const http = yield* HttpClient.HttpClient
      const findById = Effect.fn("Users.findById")(function* (id: UserId) {
        const response = yield* http.get(`/users/${id}`)
        return yield* HttpClientResponse.schemaBodyJson(User)(response)
      })
      const all = Effect.fn("Users.all")(function* () {
        const response = yield* http.get("/users")
        return yield* HttpClientResponse.schemaBodyJson(Schema.Array(User))(response)
      })
      return { findById, all }
    })
  )
}

Rules:

  • Tag identifiers must be unique. Use @app/ServiceName pattern
  • Service methods should have R = never (dependencies via Layer, not method signatures)
  • Use readonly properties

See references/services-and-layers.md for service-driven development, test layers, layer memoization, and full composition patterns.

Schema.Class and Branded Types

Use Schema.Class for domain records. Brand all entity IDs and domain primitives:

import { Schema } from "effect"

const UserId = Schema.String.pipe(Schema.brand("UserId"))
type UserId = typeof UserId.Type

const Email = Schema.String.pipe(Schema.brand("Email"))
type Email = typeof Email.Type

class User extends Schema.Class("User")({
  id: UserId,
  name: Schema.String,
  email: Email,
  createdAt: Schema.Date,
}) {
  get displayName() { return `${this.name} (${this.email})` }
}

// Construct with makeUnsafe for brands
const userId = UserId.makeUnsafe("user-123")

Use Schema.TaggedClass + Schema.Union for variants (OR types):

import { Match, Schema } from "effect"

class Success extends Schema.TaggedClass("Success")("Success", {
  value: Schema.Number,
}) {}
class Failure extends Schema.TaggedClass("Failure")("Failure", {
  error: Schema.String,
}) {}
const Result = Schema.Union([Success, Failure])
type Result = typeof Result.Type

// Exhaustive pattern matching
const render = (r: Result) => Match.valueTags(r, {
  Success: ({ value }) => `Got: ${value}`,
  Failure: ({ error }) => `Error: ${error}`,
})

See references/data-modeling.md for JSON encoding, Schema.Literals, validation, and full patterns.

Schema.TaggedErrorClass

Define domain errors with Schema.TaggedErrorClass. They are yieldable (no Effect.fail needed):

import { Schema } from "effect"

class UserNotFoundError extends Schema.TaggedErrorClass("UserNotFoundError")(
  "UserNotFoundError",
  { userId: UserId, message: Schema.String }
) {}

// Yieldable: yield directly in generators
const getUser = Effect.fn("getUser")(function* (id: UserId) {
  const user = yield* findUser(id)
  if (!user) yield* new UserNotFoundError({ userId: id, message: "Not found" })
  return user
})

Recover with catchTag / catchTags:

// Single tag
const recovered = program.pipe(
  Effect.catchTag("UserNotFoundError", (e) =>
    Effect.succeed(`User ${e.userId} missing`)
  )
)

// Multiple tags
const recovered2 = program.pipe(
  Effect.catchTags({
    UserNotFoundError: (e) => Effect.succeed("not found"),
    ValidationError: (e) => Effect.succeed("invalid"),
  })
)

See references/error-handling.md for defects, Schema.Defect, and recovery patterns.

Layer Composition

Compose layers with Layer.provideMerge (incremental, flat types) and Layer.merge (parallel):

import { Effect, Layer } from "effect"

// Compose layers for the app
const appLayer = UserService.layer.pipe(
  Layer.provideMerge(DatabaseLayer),
  Layer.provideMerge(LoggerLayer),
  Layer.provideMerge(ConfigLayer),
)

// Provide once at the entry point
const main = program.pipe(Effect.provide(appLayer))
Effect.runPromise(main)

Key rules:

  • Store parameterized layers in constants (layer memoization by reference identity)
  • Provide once at app entry, not scattered throughout code
  • Use Layer.sync for synchronous implementations, Layer.effect for effectful ones

Testing Quick Start

import { describe, expect, it } from "@effect/vitest"
import { Effect, Layer } from "effect"

it.effect("queries database", () =>
  Effect.gen(function* () {
    const db = yield* Database
    const results = yield* db.query("SELECT *")
    expect(results.length).toBe(2)
  }).pipe(Effect.provide(Database.testLayer))
)
  • Use it.effect for Effect-based tests (provides TestContext with TestClock)
  • Use it.live for real time / real clock
  • Provide fresh layers per test to prevent state leakage
  • Use it.layer only when sharing expensive resources across a suite

See references/testing.md for the full worked example and advanced patterns.

Pipe for Instrumentation

const program = fetchData.pipe(
  Effect.timeout("5 seconds"),
  Effect.retry(Schedule.exponential("100 millis").pipe(
    Schedule.compose(Schedule.recurs(3))
  )),
  Effect.tap((data) => Effect.logInfo(`Fetched: ${data}`)),
  Effect.withSpan("fetchData"),
)

Anti-Patterns

Do NotDo Instead
console.log(...)Effect.log(...) with structured data
process.env.KEYConfig.string("KEY") or Config.redacted("KEY")
throw new Error() inside Effect.genyield* new TaggedError({...}) or Effect.fail(...)
Effect.runSync(...) inside servicesKeep everything effectful
Effect.catchAll(() => ...) losing type infoEffect.catchTag / Effect.catchTags
null / undefined in domain typesOption<T> with Option.match
Option.getOrThrow(...)Option.match({ onNone, onSome }) or Option.getOrElse
Effect.Service (v3)ServiceMap.Service (v4)
Schema.TaggedError<T>() (v3)Schema.TaggedErrorClass("Tag")("Tag", {...}) (v4)
Scatter Effect.provide callsProvide once at app entry
Call parameterized layer constructors inlineStore layers in constants (memoization)

Reference Files

Load these as needed for deeper patterns:

  • Services & Layers: ServiceMap.Service, service-driven development, test layers, layer memoization, provide vs provideMerge
  • Data Modeling: Schema.Class, branded types, variants, Match.valueTags, JSON encoding
  • Schema Decisions: Schema.Class vs Struct vs TaggedClass decision flowchart, migration patterns
  • Error Handling: Schema.TaggedErrorClass, catch/catchTag/catchTags, defects, Schema.Defect, TypeId/refail patterns
  • Testing: @effect/vitest setup, it.effect/it.live/it.layer, TestClock, Effect.flip, FiberRef isolation, worked example
  • HTTP Clients: HttpClient, request building, response decoding, middleware, retries, typed API service
  • CLI: Command.make, Arguments, Flags, subcommands, worked task manager example
  • Config: Config module, schema validation, ConfigProvider, Redacted, config layers
  • Processes & Scopes: Fork types, Scope.extend, Command for child processes, killable background tasks
  • Setup: tsconfig, Effect Language Service, project structure, module settings

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.