Feature Flag System

Implement feature flags to toggle features, run A/B tests, and control rollouts without redeployment.

Purpose

Enable gradual feature rollouts, A/B testing, kill switches, and environment-specific behavior.

When to Use

• Gradual feature rollouts (10% → 50% → 100%)
• A/B testing different versions
• Beta features for select users
• Kill switches for problematic features
• Environment-specific features

Quick Start

python3 /home/ubuntu/skills/feature-flag-system/scripts/setup_flags.py /path/to/project

Usage Example

import { useFeatureFlag } from '@/lib/flags'

export function Dashboard() {
  const { enabled } = useFeatureFlag('new_dashboard')
  
  if (enabled) return <NewDashboard />
  return <OldDashboard />
}

Common Patterns

Percentage Rollout

// Roll out to 25% of users
const rollout = 25
const userHash = hashUserId(user.id) % 100
if (userHash < rollout) {
  // Show new feature
}

User-Based Access

const betaUsers = ['user1', 'user2']
if (betaUsers.includes(user.id)) {
  // Show beta feature
}

Providers

• Vercel Flags - Built-in, free
• PostHog - Open source, analytics included
• LaunchDarkly - Enterprise-grade
• Split.io - A/B testing focused

Time Savings

Manual: 1-2 hours
With Skill: 10-15 minutes
Saved: ~1.5 hours per project

Files

• scripts/setup_flags.py - Automated setup
• templates/feature-flag.tsx - React hook
• references/flags_patterns.md - Best practices

Related Skills

• analytics-dashboard - Track flag usage
• deployment-automation - Deploy with flags
• testing-framework - Test flag variations

Ship features safely with instant rollback capability.