Install a Feature Management SDK and evaluate your first flag

Install the Wingify Feature Management SDK for Node.js or Python, initialize it with your Account ID and SDK key, evaluate a feature flag for a user and track a conversion event.

#When to use this

Your product team has created a feature flag in Wingify and now a developer needs to wire it into the backend. This guide takes you from an empty project to a working getFlag() call and a tracked conversion, using the Node.js or Python SDK. The same four calls (init, user context, getFlag, trackEvent) exist in every Feature Management SDK, so the flow also applies to Java, .NET, Go, PHP and Ruby.

#Before you start

• A feature flag exists in Feature Management > Feature Flags, with at least one rule switched on in the environment you'll test. See Create a feature flag and use it in your code.
• The flag's metrics are based on events defined in Data Platform > Events. trackEvent() only records events that already exist there.
• Node.js 12.0 or later, or Python 3.6 or later.
• Your application can pass a stable, unique ID for each user.

#Steps

#1. Pick the SDK for your stack

Server-side SDKs: Node.js, Java, Python, Ruby, .NET, Go and PHP. Client-side SDKs cover Android, iOS, React Native, Flutter, React and JavaScript (web). This guide shows Node.js and Python; see the list of SDKs for the others.

#2. Get your SDK key and Account ID

Go to Configurations > Websites and Apps. On the Introduction page, find the Feature Management card ("Manage environments and view SDK keys in your pre-configured project") and click View Project. Each environment (for example, Production, Staging, Development) has its own SDK key, so copy the key for the environment you're working in. Your Account ID is shown next to the page title in Settings.

#3. Install the SDK

Node.js (npm or yarn):

npm install wingify-fme-node-sdk --save

yarn add wingify-fme-node-sdk

Python (pip):

pip install wingify-fme-python-sdk

#4. Initialize the client once, at startup

Call init() near your application's existing setup code and reuse the returned client.

Node.js:

const { init } = require('wingify-fme-node-sdk');

const wingifyClient = await init({
  accountId: '123456', // Wingify Account ID
  sdkKey: '32-alpha-numeric-sdk-key', // SDK Key
});

Python:

from wingify import init

options = {
   'sdk_key': '32-alpha-numeric-sdk-key', # SDK Key
   'account_id': '123456' # Wingify Account ID
}

wingify_client = init(options)

#5. Build the user context

The user context identifies the user. id is the only required key; add custom variables, user agent or IP address when your rules target on them.

Node.js:

const userContext = {
  id: 'unique_user_id',
  customVariables: { age: 25, location: 'US' },
};

Python:

user_context = {
    'id' : 'unique_user_id',
    'custom_variables' : { 'age' : 25, 'location' : 'US' }
}

#6. Evaluate the flag

Call getFlag() with the flag key and the user context, then branch on whether it's enabled and read variables with a default value.

Node.js:

const flag = await wingifyClient.getFlag('feature_key', userContext);
const isFlagEnabled = flag.isEnabled();
variableValue = flag.getVariable('variableKey', 'defaultValue')

Python:

flag = wingify_client.get_flag('feature_key', user_context)
is_enabled = flag.is_enabled()
variable_value = flag.get_variable('variable_key', 'default_value')

isEnabled() returns true when the user qualifies for any rollout, testing or personalize rule on the flag. The flag key is shown when you create the flag and under the flag's settings afterwards.

#7. Track a conversion

When the user completes the action your metric measures, send the event with the same user context.

Node.js:

wingifyClient.trackEvent('event-name', userContext);

Python:

wingify_client.track_event('event_name', user_context)

You can pass event properties as a third argument, for example { userType: "free" }. Property names must match the ones defined on the event in Data Platform > Events.

#Check that it worked

• Call getFlag() with a test user who meets your rule and confirm isEnabled() returns true; switch the rule off in Wingify and confirm it returns false once settings refresh.
• Set the logger level to DEBUG in init() to see SDK decisions in your server console.
• In Configurations > Websites and Apps, open your Feature Management project and use Logs for advanced debugging.

#Common questions

getFlag() always returns false. Why? The flag must already exist in Wingify and have a rule that the user qualifies for in the environment whose SDK key you used. Check the key, the environment and the rule status.

Can I segment reports on user context attributes? No. User context attributes are used for targeting. For post-segmentation in reports, send attributes with setAttribute() instead.

Is it safe to expose the SDK key? The SDK key is read-only: it can fetch flag configurations but can't change them. For browser or mobile use, create a dedicated environment that contains only client-safe flags.

I'm migrating from FullStack. What changed? FullStack used getSettingsFile() + launch(), activate() and campaign-based track(). Feature Management uses init(), getFlag() and trackEvent(). See Work with FullStack projects.

#Learn more

• List of Feature Management SDKs
• Node.js SDK: Initialization
• Python SDK: Initialization
• SDK Keys in Client-Side Environments
• Getting Started with Wingify Feature Management