Offers a vendor-agnostic API for feature flagging, enabling integration with various feature management tools. Supports targeting, hooks, providers, and tracking for contextually-aware flag evaluation.
OpenFeature is an open specification that provides a vendor-agnostic, community-driven API for feature flagging that works with your favorite feature flag management tool.
The following Kotlin Multiplatform Targets are supported:
Note that this library adheres to the Static Context Paradigm, so it is intended to be used on the client side (i.e. mobile apps, web apps and desktop apps), and has not been evaluated for use in other types of applications (e.g. server applications).
Installation is preferred via Maven Central.
[!IMPORTANT] Before version 0.6.0 the Maven artifact's id was
android-sdk. When upgrading to 0.6.0 or higher please do make sure to use the new artifact id, which iskotlin-sdk.
dependencies {
api("dev.openfeature:kotlin-sdk:0.8.0")
}
kotlin {
sourceSets {
commonMain.dependencies {
api("dev.openfeature:kotlin-sdk:0.8.0")
}
}
}
[!NOTE] In version 0.6.0 the base package name has changed from
dev.openfeature.sdktodev.openfeature.kotlin.sdk. When upgrading to 0.6.0 or higher please update your imports accordingly.Example:
import dev.openfeature.sdk.EvaluationContext->import dev.openfeature.kotlin.sdk.EvaluationContext.
coroutineScope.launch(Dispatchers.Default) {
// configure a provider, wait for it to complete its initialization tasks
OpenFeatureAPI.setProviderAndWait(customProvider)
val client = OpenFeatureAPI.getClient()
// get a bool flag value
client.getBooleanValue("boolFlag", defaultValue = false)
}
Implemented: ✅ | In-progress: ⚠️ | Not implemented yet: ❌
Providers are an abstraction between a flag management system and the OpenFeature SDK. Look here for a complete list of available providers. If the provider you're looking for hasn't been created yet, see the develop a provider section to learn how to build it yourself.
Once you've added a provider as a dependency, it can be registered with OpenFeature like this:
coroutineScope.launch(Dispatchers.Default) {
OpenFeatureAPI.setProviderAndWait(MyProvider())
}
Asynchronous API that doesn't wait is also available. It's useful when you want to set a provider and continue with other tasks.
However, flag evaluations are only possible after the provider is Ready.
OpenFeatureAPI.setProvider(MyProvider()) // can pass a dispatcher here
// The provider initialization happens on a coroutine launched on the IO dispatcher.
val status = OpenFeatureAPI.getStatus()
// When status is Ready, flag evaluations can be made
Sometimes, the value of a flag must consider some dynamic criteria about the application or user, such as the user's location, IP, email address, or the server's location. In OpenFeature, we refer to this as targeting. If the flag management system you're using supports targeting, you can provide the input data using the evaluation context.
// set a value to the global context
val evaluationContext = ImmutableContext(
targetingKey = session.getId,
attributes = mutableMapOf("region" to Value.String("us-east-1")))
OpenFeatureAPI.setEvaluationContext(evaluationContext)
Hooks allow for custom logic to be added at well-defined points of the flag evaluation life-cycle. Look here for a complete list of available hooks. If the hook you're looking for hasn't been created yet, see the develop a hook section to learn how to build it yourself.
Once you've added a hook as a dependency, it can be registered at the global, client, or flag invocation level.
// add a hook globally, to run on all evaluations
OpenFeatureAPI.addHooks(listOf(ExampleHook()))
// add a hook on this client, to run on all evaluations made by this client
val client = OpenFeatureAPI.getClient()
client.addHooks(listOf(ExampleHook()))
// add a hook for this evaluation only
val retval = client.getBooleanValue(flagKey, false,
FlagEvaluationOptions(listOf(ExampleHook())))
The tracking API allows you to use
OpenFeature abstractions to associate user actions with feature flag evaluations.
This is essential for robust experimentation powered by feature flags. Note that, unlike methods
that handle feature flag evaluations, calling track(...) may throw an IllegalArgumentException
if an empty string is passed as the trackingEventName.
Below is an example of how we can track a "Checkout" event with some TrackingDetails.
OpenFeatureAPI.getClient().track(
"Checkout",
TrackingEventDetails(
499.99,
ImmutableStructure(
"numberOfItems" to Value.Integer(4),
"timeInCheckout" to Value.String("PT3M20S")
)
)
)
Tracking is optionally implemented by Providers.
The SDK ships a Logger interface and a built-in LoggingHook that emits structured log records at each stage of the flag evaluation life-cycle.
Both message and attributes are lambdas — they are evaluated lazily and only invoked if the logger decides to emit the record. This means inactive log levels incur no allocation overhead.
Platform-specific loggers are created via LoggerFactory.getLogger(tag):
To route SDK logs into your own logging framework, implement Logger and pass it wherever a Logger is accepted:
LoggingHook logs at each stage of flag evaluation. Register it like any other hook:
// dev.openfeature.kotlin.sdk.logging.LoggerFactory — the SDK's platform-specific factory
val logger = LoggerFactory.getLogger("FeatureFlags")
val hook = LoggingHook(
logger = logger,
logEvaluationContext = false, // set true to include context attributes in logs
beforeLogLevel = LogLevel.DEBUG,
afterLogLevel = LogLevel.DEBUG,
errorLogLevel = LogLevel.ERROR,
finallyLogLevel = LogLevel.DEBUG,
)
// register globally
OpenFeatureAPI.addHooks(listOf(hook))
When logEvaluationContext = true, context attributes are included in log records. By default, attributes matching common PII field names (see LoggingHook.DEFAULT_SENSITIVE_KEYS) are excluded. Two optional parameters let you customize this behaviour:
contains: , , , , , , , , , , , , , , , , .
// Log only a specific set of safe attributes
val hook = LoggingHook(
logger = logger,
logEvaluationContext = true,
logTargetingKey = false, // targeting key contains a user ID
includeAttributes = setOf("region", "plan", "tier"), // allowlist — only these are logged
)
// Or extend the default denylist
val hook = LoggingHook(
logger = logger,
logEvaluationContext = true,
excludeAttributes = LoggingHook.DEFAULT_SENSITIVE_KEYS + setOf(, ),
)
The hook logs the following at each lifecycle stage:
Example JVM output for a successful evaluation:
2026-04-15T10:00:00.123Z [DEBUG] FeatureFlags - Flag evaluation starting flag=my-flag type=BOOLEAN defaultValue=false provider=MyProvider
2026-04-15T10:00:00.124Z [DEBUG] FeatureFlags - Flag evaluation completed flag=my-flag value=true variant=on reason=TARGETING_MATCH provider=MyProvider
2026-04-15T10:00:00.124Z [DEBUG] FeatureFlags - Flag evaluation finalized flag=my-flag
To include evaluation context in logs for a single call, pass a hook hint:
client.getBooleanValue(
"my-flag",
false,
FlagEvaluationOptions(
hooks = listOf(hook),
hookHints = mapOf(LoggingHook.HINT_LOG_EVALUATION_CONTEXT to true)
)
)
Domains allow you to logically bind clients with providers, enabling multiple provider configurations within the same application. Support for domains is currently in development.
Events from the Provider allow the SDK to react to state changes in the provider or underlying flag management system, such as flag definition changes, provider readiness, or error conditions.
Events are optional which mean that not all Providers will emit them and it is not a must have. Some providers support additional events, such as PROVIDER_CONFIGURATION_CHANGED.
Please refer to the documentation of the provider you're using to see what events are supported.
Example usage:
viewModelScope.launch {
OpenFeatureAPI.observe<OpenFeatureProviderEvents>().collect {
println(">> Provider event received")
}
}
viewModelScope.launch {
OpenFeatureAPI.setProviderAndWait(
MyFeatureProvider(),
myEvaluationContext,
Dispatchers.Default
)
}
The Multi-Provider allows you to use multiple underlying providers as sources of flag data for the OpenFeature Kotlin SDK. When a flag is being evaluated, the Multi-Provider will use each underlying provider in order to determine the final result. Different evaluation strategies can be defined to control which providers get evaluated and which result is used.
The Multi-Provider is a useful tool for performing migrations between flag providers, or combining multiple providers into a single feature flagging interface.
See the Multi-Provider documentation for more details.
The OpenFeature API provides a close function to perform a cleanup of the registered provider. This should only be called when your application is in the process of shutting down.
coroutineScope.launch {
OpenFeatureAPI.shutdown()
}
In the repo there is also a sample app currently under development. The sample app can be used to try out development of a Provider, a Hook or to validate changes to the SDK itself.
The sample app should not be used as a reference implementation of how to use the OpenFeature SDK in an Android app.
To develop a provider, you need to create a new project and include the OpenFeature SDK as a dependency.
You’ll then need to write the provider by implementing the FeatureProvider interface exported by the OpenFeature SDK.
Built a new provider? Let us know so we can add it to the docs!
To develop a hook, you need to create a new project and include the OpenFeature SDK as a dependency.
Implement your own hook by conforming to the Hook interface exported by the OpenFeature SDK.
Built a new hook? Let us know so we can add it to the docs!
SNAPSHOT versions are published to Maven Central's snapshot repository from the upcoming release branch. If you need a SNAPSHOT version published, please reach out to the maintainers to trigger the workflow.
Interested in contributing? Great, we'd love your help! To get started, take a look at the CONTRIBUTING guide.
Made with contrib.rocks.
| Supported | Platform | Supported versions |
|---|
| ✅ | Android | SDK 21+ |
| ✅ | JVM | JDK 11+ |
| ✅ | Native | Linux x64, iosSimulatorArm64, iosArm64 |
| ❌ | Native | Other native targets |
| ✅ | Javascript (Node.js) | |
| ✅ | Javascript (Browser) | |
| ❌ | Wasm |
| Status | Features | Description |
|---|
| ✅ | Providers | Integrate with a commercial, open source, or in-house feature management tool. |
| ✅ | Targeting | Contextually-aware flag evaluation using evaluation context. |
| ✅ | Hooks | Add functionality to various stages of the flag evaluation life-cycle. |
| ✅ | Tracking | Associate user actions with feature flag evaluations. |
| ✅ | Logging | Integrate with popular logging packages. |
| ⚠️ | Domains | Logically bind clients with providers. |
| ✅ | Eventing | React to state changes in the provider or flag management system. |
| ✅ | Shutdown | Gracefully clean up a provider during application shutdown. |
| ✅ | Extending | Extend OpenFeature with custom providers and hooks. |
| ✅ | Multi-Provider | Combine multiple providers with configurable evaluation strategies. |
interface Logger {
fun debug(message: () -> String, attributes: () -> Map<String, Any?> = { emptyMap() }, throwable: Throwable? = null)
fun info(message: () -> String, attributes: () -> Map<String, Any?> = { emptyMap() }, throwable: Throwable? = null)
fun warn(message: () -> String, attributes: () -> Map<String, Any?> = { emptyMap() }, throwable: Throwable? = null)
fun error(message: () -> String, attributes: () -> Map<String, Any?> = { emptyMap() }, throwable: Throwable? = null)
}
| Platform | Backend | Format |
|---|
| Android | android.util.Log | Logcat with tag; attributes appended as key=value pairs |
| JVM | System.out (DEBUG/INFO) / System.err (WARN/ERROR) | <timestamp> [LEVEL] <tag> - <message> key=value … |
| iOS | NSLog | [LEVEL] <tag> - <message> key=value … (NSLog adds its own timestamp) |
| JavaScript | console API | [<tag>] <message> with attributes as an expandable JS object (browser devtools / Node.js); note: debug uses console.log, not console.debug — browser "Verbose" filter will not capture it |
| Linux/Native | stdout (DEBUG/INFO) / stderr (WARN/ERROR) | [LEVEL] <tag> - <message> key=value … (no timestamp; systemd/journald provides its own) |
// SLF4J's {} placeholder calls toString() on the map, producing {key=value, ...}.
// Use entries.joinToString(" ") { "${it.key}=${it.value}" } to get key=value output instead.
class MyLogger(private val underlying: org.slf4j.Logger) : Logger {
override fun debug(message: () -> String, attributes: () -> Map<String, Any?>, throwable: Throwable?) {
if (!underlying.isDebugEnabled) return
underlying.debug("{} {}", message(), attributes(), throwable)
}
override fun info(message: () -> String, attributes: () -> Map<String, Any?>, throwable: Throwable?) {
if (!underlying.isInfoEnabled) return
underlying.info("{} {}", message(), attributes(), throwable)
}
override fun warn(message: () -> String, attributes: () -> Map<String, Any?>, throwable: Throwable?) {
if (!underlying.isWarnEnabled) return
underlying.warn("{} {}", message(), attributes(), throwable)
}
override fun error(message: () -> String, attributes: () -> Map<String, Any?>, throwable: Throwable?) {
if (!underlying.isErrorEnabled) return
underlying.error("{} {}", message(), attributes(), throwable)
}
}
| Parameter | Type | Default | Description |
|---|
logTargetingKey | Boolean | true | Include the targeting key in context logs. Set to false if targeting keys contain PII such as user IDs or email addresses. |
includeAttributes | Set<String>? | null | Allowlist: only attributes with these names are logged. An empty set logs no attributes. Takes precedence over excludeAttributes. Attribute name matching is case-sensitive. |
excludeAttributes | Set<String> | DEFAULT_SENSITIVE_KEYS | Denylist: attributes with these names are omitted. Attribute name matching is case-sensitive. |
LoggingHook.DEFAULT_SENSITIVE_KEYSemailphonephoneNumberssnsocialSecurityNumbercreditCardcreditCardNumberpasswordaddressstreetAddresszipCodepostalCodeipAddressfirstNamelastNamefullNamedateOfBirth| Stage | Message | Attributes |
|---|
before | Flag evaluation starting | flag, type, defaultValue, provider, client (if set), context.* (if enabled) |
after | Flag evaluation completed | flag, value, variant (if set), reason (if set), provider, context.* (if enabled) |
error | Flag evaluation error | flag, type, defaultValue, provider, error (if set), context.* (if enabled) |
finallyAfter | Flag evaluation finalized | flag, errorCode (if set), errorMessage (if set) |
class NewProvider(override val hooks: List<Hook<*>>, override val metadata: ProviderMetadata) : FeatureProvider {
override fun getBooleanEvaluation(
key: String,
defaultValue: Boolean,
context: EvaluationContext?
): ProviderEvaluation<Boolean> {
// resolve a boolean flag value
}
override fun getDoubleEvaluation(
key: String,
defaultValue: Double,
context: EvaluationContext?
): ProviderEvaluation<Double> {
// resolve a double flag value
}
override fun getIntegerEvaluation(
key: String,
defaultValue: Int,
context: EvaluationContext?
): ProviderEvaluation<Int> {
// resolve an integer flag value
}
override fun getLongEvaluation(
key: String,
defaultValue: Long,
context: EvaluationContext?
): ProviderEvaluation<Long> {
// resolve a long (64-bit integer) flag value
}
override fun getObjectEvaluation(
key: String,
defaultValue: Value,
context: EvaluationContext?
): ProviderEvaluation<Value> {
// resolve an object flag value
}
override fun getStringEvaluation(
key: String,
defaultValue: String,
context: EvaluationContext?
): ProviderEvaluation<String> {
// resolve a string flag value
}
override suspend fun initialize(initialContext: EvaluationContext?) {
// add context-aware provider initialization
}
override suspend fun onContextSet(oldContext: EvaluationContext?, newContext: EvaluationContext) {
// add necessary changes on context change
}
override fun track(
trackingEventName: String,
context: EvaluationContext?,
details: TrackingEventDetails?
) {
// Optionally track an event
}
override fun observe(): Flow<OpenFeatureProviderEvents> {
// Optionally return a `Flow` of OpenFeatureProviderEvents
}
}
Surfaced from shared tags and platforms — no rankings paid for.