IdGenerator
Provides identifier generation for AI features.
The IdGenerator service exposes one operation, generateId, which returns
a string inside Effect. AI modules use it for values such as tool call ids
and generated response item ids. This module includes the service tag,
service interface, default generator, configurable custom generator, and layer
for providing the service.
Constructors
defaultIdGenerator
Default ID generator service implementation.
Details
Uses the standard configuration with "id" prefix and generates IDs in the format "id_XXXXXXXXXXXXXXXX" where X represents random alphanumeric characters.
Signature
declare const defaultIdGenerator: ServiceExample
(Generating default IDs)
import { Effect } from "effect"import { IdGenerator } from "effect/unstable/ai"
const program = Effect.gen(function*() { const id = yield* IdGenerator.defaultIdGenerator.generateId() return id})
// Or provide it as a serviceconst withDefault = program.pipe( Effect.provideService( IdGenerator.IdGenerator, IdGenerator.defaultIdGenerator ))
const id = await Effect.runPromise(withDefault)const result = [id.startsWith("id_"), id.length] // => [true, 19]Creates a custom ID generator service with the specified options.
Details
Validates the configuration to ensure the separator is not part of the alphabet, which would cause ambiguity in parsing generated IDs.
Signature
declare const make: (...args: [MakeOptions]) => Effect<{ generateId: (...args: []) => Effect<string, never, never>;}, IllegalArgumentError, never>Example
(Creating a custom generator)
import { Effect } from "effect"import { IdGenerator } from "effect/unstable/ai"
const program = Effect.gen(function*() { // Create a generator for AI assistant message IDs const messageIdGen = yield* IdGenerator.make({ alphabet: "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", prefix: "msg", separator: "-", size: 10 })
return yield* messageIdGen.generateId()})
const messageId = await Effect.runPromise(program)const result = [messageId.startsWith("msg-"), messageId.length] // => [true, 14]Example
(Handling invalid generator options)
import { Effect } from "effect"import { IdGenerator } from "effect/unstable/ai"
// This will fail with IllegalArgumentErrorconst invalidConfig = IdGenerator.make({ alphabet: "ABC123", prefix: "test", separator: "A", // Error: separator is part of alphabet size: 8})
const error = await Effect.runPromise(Effect.flip(invalidConfig))error.message // => 'The separator "A" must not be part of the alphabet "ABC123".'Layers
Creates a Layer that provides the IdGenerator service with custom configuration.
When to use
Use when you need to provide ID generation capabilities from validated configuration.
Signature
declare function layer(options: MakeOptions): Layer<IdGenerator, IllegalArgumentError>Example
(Providing an ID generator layer)
import { Effect } from "effect"import { IdGenerator } from "effect/unstable/ai"
// Create a layer for generating AI tool call IDsconst toolCallIdLayer = IdGenerator.layer({ alphabet: "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ", prefix: "tool_call", separator: "_", size: 12})
const program = Effect.gen(function*() { const idGen = yield* IdGenerator.IdGenerator return yield* idGen.generateId()}).pipe(Effect.provide(toolCallIdLayer))
const toolCallId = await Effect.runPromise(program)const result = [toolCallId.startsWith("tool_call_"), toolCallId.length] // => [true, 22]Models
The service interface for ID generation.
Details
Defines the contract that all ID generator implementations must fulfill. The service provides a single method for generating unique identifiers in an effectful context.
Signature
interface Service { readonly generateId: () => Effect<string>;}Example
(Implementing a custom ID generator)
import { Effect } from "effect"import type { IdGenerator } from "effect/unstable/ai"
// Custom deterministic implementationlet nextId = 0const customService: IdGenerator.Service = { generateId: () => Effect.sync(() => `custom_${++nextId}`)}
const program = customService.generateId()
await Effect.runPromise(program) // => "custom_1"Options
MakeOptions interface
Configuration options for creating custom ID generators.
Signature
interface MakeOptions { readonly alphabet: string; readonly prefix?: string; readonly separator: string; readonly size: number;}Example
(Configuring generated IDs)
import type { IdGenerator } from "effect/unstable/ai"
// Configuration for tool call IDsconst toolCallOptions: IdGenerator.MakeOptions = { alphabet: "0123456789ABCDEF", prefix: "tool", separator: "_", size: 8}
// This will generate IDs like: "tool_A1B2C3D4"const result = [toolCallOptions.prefix, toolCallOptions.size] // => ["tool", 8]Services
IdGenerator
Service tag for AI identifier generation services.
When to use
Use to access or provide the service that creates identifiers for AI tool calls and related generated values.
Details
This tag is used to provide and access ID generation functionality throughout the application. It follows Effect's standard service pattern for type-safe dependency injection.
Signature
declare class IdGenerator extends Shape<"@effect/ai/IdGenerator", Service, this> { constructor(_: never);}Example
(Accessing the ID generator service)
import { Effect } from "effect"import { IdGenerator } from "effect/unstable/ai"
const useIdGenerator = Effect.gen(function*() { const idGenerator = yield* IdGenerator.IdGenerator const newId = yield* idGenerator.generateId() return newId})
const program = useIdGenerator.pipe( Effect.provideService(IdGenerator.IdGenerator, { generateId: () => Effect.succeed("id-1") }))await Effect.runPromise(program) // => "id-1"