Three simple steps to visualize it:

1. Manage your Featurevisor project in a Git repository
2. Build and upload datafiles (static JSON files) to your CDN or custom server
3. Fetch the datafile in your application, and evaluate features using SDKs

• A Git repository for managing your project declaratively
• A CI/CD pipeline (like GitHub Actions) for building and uploading datafiles
• A CDN or custom server for serving the generated datafiles

Featurevisor SDKs will take care of the rest for you.

Three core building blocks of a Featurevisor project are:

• Attributes: building block for conditions
• Segments: reusable conditions for targeting users
• Features: features with flag, variation, and variable definitions

SDKs are used in your application(s) to evaluate three different types of values belonging to your defined features against your provided context in the runtime:

• Flags: on/off values against rules
• Variations: string values for A/B tests
• Variables: key/value pairs for remote configuration

Supported SDKs include:

• JavaScript (both Node.js and browser environments)
• React
• React Native
• Go
• Python
• PHP
• Ruby
• Java
• Swift
• Next.js
• Vue.js
• OpenFeature providers for Node.js, browsers, and other supported SDK languages

Featurevisor ships an official skill that teaches AI agents (like Claude Code, Cursor, Codex, and OpenCode) how to work with your project: creating features and segments, writing rollout rules and test specs, debugging evaluations, promoting between sets, and integrating SDKs in your applications.

Install it in your Featurevisor project (or your app repository) via skills.sh:

$ npx skills add featurevisor/featurevisor

Then prompt your agent naturally:

• "Create a showWishlist feature, tagged web, starting at 5% in production"
• "Why is checkout disabled for userId 123 in the Netherlands?"
• "Roll out newEditor to employees first, then beta users, then everyone"

Pair it with npx featurevisor catalog running locally to visually review every change your agent makes, live in the browser.

The skill's source lives in this repository under skills/.

Several use cases include:

• Decoupling application deployments from feature releases
• Progressive delivery with gradual rollout of features
• Testing in production
• Experiment with A/B and Multivariate tests
• Role-based feature gating
• Remote configuration with advanced variables
• Establishing feature ownership and governance
• Trunk based development
• Aligning multiple teams in a microfrontends architecture
• Feature depdendencies management
• Deprecating features safely

MIT © Fahad Heylaal